Merge pull request #485 from kunish/fix/copilot-premium-request-inflation

fix(copilot): reduce premium request inflation, enable thinking, and use dynamic API limits

.gitignore

@@ -37,13 +37,13 @@ GEMINI.md
 
 # Tooling metadata
 .vscode/*
+.worktrees/
 .codex/*
 .claude/*
 .gemini/*
 .serena/*
 .agent/*
 .agents/*
-.agents/*
 .opencode/*
 .idea/*
 .bmad/*

README_CN.md

@@ -1,6 +1,6 @@
 # CLIProxyAPI Plus
 
-[English](README.md) | 中文 | [日本語](README_JA.md)
+[English](README.md) | 中文
 
 这是 [CLIProxyAPI](https://github.com/router-for-me/CLIProxyAPI) 的 Plus 版本，在原有基础上增加了第三方供应商的支持。
 

README_JA.md

@@ -1,187 +0,0 @@
-# CLI Proxy API
-
-[English](README.md) | [中文](README_CN.md) | 日本語
-
-CLI向けのOpenAI/Gemini/Claude/Codex互換APIインターフェースを提供するプロキシサーバーです。
-
-OAuth経由でOpenAI Codex（GPTモデル）およびClaude Codeもサポートしています。
-
-ローカルまたはマルチアカウントのCLIアクセスを、OpenAI（Responses含む）/Gemini/Claude互換のクライアントやSDKで利用できます。
-
-## スポンサー
-
-[![z.ai](https://assets.router-for.me/english-5-0.jpg)](https://z.ai/subscribe?ic=8JVLJQFSKB)
-
-本プロジェクトはZ.aiにスポンサーされており、GLM CODING PLANの提供を受けています。
-
-GLM CODING PLANはAIコーディング向けに設計されたサブスクリプションサービスで、月額わずか$10から利用可能です。フラッグシップのGLM-4.7および（GLM-5はProユーザーのみ利用可能）モデルを10以上の人気AIコーディングツール（Claude Code、Cline、Roo Codeなど）で利用でき、開発者にトップクラスの高速かつ安定したコーディング体験を提供します。
-
-GLM CODING PLANを10%割引で取得：https://z.ai/subscribe?ic=8JVLJQFSKB
-
----
-
-<table>
-<tbody>
-<tr>
-<td width="180"><a href="https://www.packyapi.com/register?aff=cliproxyapi"><img src="./assets/packycode.png" alt="PackyCode" width="150"></a></td>
-<td>PackyCodeのスポンサーシップに感謝します！PackyCodeは信頼性が高く効率的なAPIリレーサービスプロバイダーで、Claude Code、Codex、Geminiなどのリレーサービスを提供しています。PackyCodeは当ソフトウェアのユーザーに特別割引を提供しています：<a href="https://www.packyapi.com/register?aff=cliproxyapi">こちらのリンク</a>から登録し、チャージ時にプロモーションコード「cliproxyapi」を入力すると10%割引になります。</td>
-</tr>
-<tr>
-<td width="180"><a href="https://www.aicodemirror.com/register?invitecode=TJNAIF"><img src="./assets/aicodemirror.png" alt="AICodeMirror" width="150"></a></td>
-<td>AICodeMirrorのスポンサーシップに感謝します！AICodeMirrorはClaude Code / Codex / Gemini CLI向けの公式高安定性リレーサービスを提供しており、エンタープライズグレードの同時接続、迅速な請求書発行、24時間365日の専任技術サポートを備えています。Claude Code / Codex / Geminiの公式チャネルが元の価格の38% / 2% / 9%で利用でき、チャージ時にはさらに割引があります！CLIProxyAPIユーザー向けの特別特典：<a href="https://www.aicodemirror.com/register?invitecode=TJNAIF">こちらのリンク</a>から登録すると、初回チャージが20%割引になり、エンタープライズのお客様は最大25%割引を受けられます！</td>
-</tr>
-<tr>
-<td width="180"><a href="https://shop.bmoplus.com/?utm_source=github"><img src="./assets/bmoplus.png" alt="BmoPlus" width="150"></a></td>
-<td>本プロジェクトにご支援いただいた BmoPlus に感謝いたします！BmoPlusは、AIサブスクリプションのヘビーユーザー向けに特化した信頼性の高いAIアカウントサービスプロバイダーであり、安定した ChatGPT Plus / ChatGPT Pro (完全保証) / Claude Pro / Super Grok / Gemini Pro の公式代行チャージおよび即納アカウントを提供しています。こちらの<a href="https://shop.bmoplus.com/?utm_source=github">BmoPlus AIアカウント専門店/代行チャージ</a>経由でご登録・ご注文いただいたユーザー様は、GPTを <b>公式サイト価格の約1割（90% OFF）</b> という驚異的な価格でご利用いただけます！</td>
-</tr>
-</tbody>
-</table>
-
-## 概要
-
-- CLIモデル向けのOpenAI/Gemini/Claude互換APIエンドポイント
-- OAuthログインによるOpenAI Codexサポート（GPTモデル）
-- OAuthログインによるClaude Codeサポート
-- OAuthログインによるQwen Codeサポート
-- OAuthログインによるiFlowサポート
-- プロバイダールーティングによるAmp CLIおよびIDE拡張機能のサポート
-- ストリーミングおよび非ストリーミングレスポンス
-- 関数呼び出し/ツールのサポート
-- マルチモーダル入力サポート（テキストと画像）
-- ラウンドロビン負荷分散による複数アカウント対応（Gemini、OpenAI、Claude、QwenおよびiFlow）
-- シンプルなCLI認証フロー（Gemini、OpenAI、Claude、QwenおよびiFlow）
-- Generative Language APIキーのサポート
-- AI Studioビルドのマルチアカウント負荷分散
-- Gemini CLIのマルチアカウント負荷分散
-- Claude Codeのマルチアカウント負荷分散
-- Qwen Codeのマルチアカウント負荷分散
-- iFlowのマルチアカウント負荷分散
-- OpenAI Codexのマルチアカウント負荷分散
-- 設定によるOpenAI互換アップストリームプロバイダー（例：OpenRouter）
-- プロキシ埋め込み用の再利用可能なGo SDK（`docs/sdk-usage.md`を参照）
-
-## はじめに
-
-CLIProxyAPIガイド：[https://help.router-for.me/](https://help.router-for.me/)
-
-## 管理API
-
-[MANAGEMENT_API.md](https://help.router-for.me/management/api)を参照
-
-## Amp CLIサポート
-
-CLIProxyAPIは[Amp CLI](https://ampcode.com)およびAmp IDE拡張機能の統合サポートを含んでおり、Google/ChatGPT/ClaudeのOAuthサブスクリプションをAmpのコーディングツールで使用できます：
-
-- Ampの APIパターン用のプロバイダールートエイリアス（`/api/provider/{provider}/v1...`）
-- OAuth認証およびアカウント機能用の管理プロキシ
-- 自動ルーティングによるスマートモデルフォールバック
-- 利用できないモデルを代替モデルにルーティングする**モデルマッピング**（例：`claude-opus-4.5` → `claude-sonnet-4`）
-- localhostのみの管理エンドポイントによるセキュリティファーストの設計
-
-**→ [Amp CLI統合ガイドの完全版](https://help.router-for.me/agent-client/amp-cli.html)**
-
-## SDKドキュメント
-
-- 使い方：[docs/sdk-usage.md](docs/sdk-usage.md)
-- 上級（エグゼキューターとトランスレーター）：[docs/sdk-advanced.md](docs/sdk-advanced.md)
-- アクセス：[docs/sdk-access.md](docs/sdk-access.md)
-- ウォッチャー：[docs/sdk-watcher.md](docs/sdk-watcher.md)
-- カスタムプロバイダーの例：`examples/custom-provider`
-
-## コントリビューション
-
-コントリビューションを歓迎します！お気軽にPull Requestを送ってください。
-
-1. リポジトリをフォーク
-2. フィーチャーブランチを作成（`git checkout -b feature/amazing-feature`）
-3. 変更をコミット（`git commit -m 'Add some amazing feature'`）
-4. ブランチにプッシュ（`git push origin feature/amazing-feature`）
-5. Pull Requestを作成
-
-## 関連プロジェクト
-
-CLIProxyAPIをベースにした以下のプロジェクトがあります：
-
-### [vibeproxy](https://github.com/automazeio/vibeproxy)
-
-macOSネイティブのメニューバーアプリで、Claude CodeとChatGPTのサブスクリプションをAIコーディングツールで使用可能 - APIキー不要
-
-### [Subtitle Translator](https://github.com/VjayC/SRT-Subtitle-Translator-Validator)
-
-CLIProxyAPI経由でGeminiサブスクリプションを使用してSRT字幕を翻訳するブラウザベースのツール。自動検証/エラー修正機能付き - APIキー不要
-
-### [CCS (Claude Code Switch)](https://github.com/kaitranntt/ccs)
-
-CLIProxyAPI OAuthを使用して複数のClaudeアカウントや代替モデル（Gemini、Codex、Antigravity）を即座に切り替えるCLIラッパー - APIキー不要
-
-### [ProxyPal](https://github.com/heyhuynhgiabuu/proxypal)
-
-CLIProxyAPI管理用のmacOSネイティブGUI：OAuth経由でプロバイダー、モデルマッピング、エンドポイントを設定 - APIキー不要
-
-### [Quotio](https://github.com/nguyenphutrong/quotio)
-
-Claude、Gemini、OpenAI、Qwen、Antigravityのサブスクリプションを統合し、リアルタイムのクォータ追跡とスマート自動フェイルオーバーを備えたmacOSネイティブのメニューバーアプリ。Claude Code、OpenCode、Droidなどのコーディングツール向け - APIキー不要
-
-### [CodMate](https://github.com/loocor/CodMate)
-
-CLI AIセッション（Codex、Claude Code、Gemini CLI）を管理するmacOS SwiftUIネイティブアプリ。統合プロバイダー管理、Gitレビュー、プロジェクト整理、グローバル検索、ターミナル統合機能を搭載。CLIProxyAPIと統合し、Codex、Claude、Gemini、Antigravity、Qwen CodeのOAuth認証を提供。単一のプロキシエンドポイントを通じた組み込みおよびサードパーティプロバイダーの再ルーティングに対応 - OAuthプロバイダーではAPIキー不要
-
-### [ProxyPilot](https://github.com/Finesssee/ProxyPilot)
-
-TUI、システムトレイ、マルチプロバイダーOAuthを備えたWindows向けCLIProxyAPIフォーク - AIコーディングツール用、APIキー不要
-
-### [Claude Proxy VSCode](https://github.com/uzhao/claude-proxy-vscode)
-
-Claude Codeモデルを素早く切り替えるVSCode拡張機能。バックエンドとしてCLIProxyAPIを統合し、バックグラウンドでの自動ライフサイクル管理を搭載
-
-### [ZeroLimit](https://github.com/0xtbug/zero-limit)
-
-CLIProxyAPIを使用してAIコーディングアシスタントのクォータを監視するTauri + React製のWindowsデスクトップアプリ。Gemini、Claude、OpenAI Codex、Antigravityアカウントの使用量をリアルタイムダッシュボード、システムトレイ統合、ワンクリックプロキシコントロールで追跡 - APIキー不要
-
-### [CPA-XXX Panel](https://github.com/ferretgeek/CPA-X)
-
-CLIProxyAPI向けの軽量Web管理パネル。ヘルスチェック、リソース監視、リアルタイムログ、自動更新、リクエスト統計、料金表示機能を搭載。ワンクリックインストールとsystemdサービスに対応
-
-### [CLIProxyAPI Tray](https://github.com/kitephp/CLIProxyAPI_Tray)
-
-PowerShellスクリプトで実装されたWindowsトレイアプリケーション。サードパーティライブラリに依存せず、ショートカットの自動作成、サイレント実行、パスワード管理、チャネル切り替え（Main / Plus）、自動ダウンロードおよび自動更新に対応
-
-### [霖君](https://github.com/wangdabaoqq/LinJun)
-
-霖君はAIプログラミングアシスタントを管理するクロスプラットフォームデスクトップアプリケーションで、macOS、Windows、Linuxシステムに対応。Claude Code、Gemini CLI、OpenAI Codex、Qwen Codeなどのコーディングツールを統合管理し、ローカルプロキシによるマルチアカウントクォータ追跡とワンクリック設定が可能
-
-### [CLIProxyAPI Dashboard](https://github.com/itsmylife44/cliproxyapi-dashboard)
-
-Next.js、React、PostgreSQLで構築されたCLIProxyAPI用のモダンなWebベース管理ダッシュボード。リアルタイムログストリーミング、構造化された設定編集、APIキー管理、Claude/Gemini/Codex向けOAuthプロバイダー統合、使用量分析、コンテナ管理、コンパニオンプラグインによるOpenCodeとの設定同期機能を搭載 - 手動でのYAML編集は不要
-
-### [All API Hub](https://github.com/qixing-jk/all-api-hub)
-
-New API互換リレーサイトアカウントをワンストップで管理するブラウザ拡張機能。残高と使用量のダッシュボード、自動チェックイン、一般的なアプリへのワンクリックキーエクスポート、ページ内API可用性テスト、チャネル/モデルの同期とリダイレクト機能を搭載。Management APIを通じてCLIProxyAPIと統合し、ワンクリックでプロバイダーのインポートと設定同期が可能
-
-### [Shadow AI](https://github.com/HEUDavid/shadow-ai)
-
-Shadow AIは制限された環境向けに特別に設計されたAIアシスタントツールです。ウィンドウや痕跡のないステルス動作モードを提供し、LAN（ローカルエリアネットワーク）を介したクロスデバイスAI質疑応答のインタラクションと制御を可能にします。本質的には「画面/音声キャプチャ + AI推論 + 低摩擦デリバリー」の自動化コラボレーションレイヤーであり、制御されたデバイスや制限された環境でアプリケーション横断的にAIアシスタントを没入的に使用できるようユーザーを支援します。
-
-> [!NOTE]
-> CLIProxyAPIをベースにプロジェクトを開発した場合は、PRを送ってこのリストに追加してください。
-
-## その他の選択肢
-
-以下のプロジェクトはCLIProxyAPIの移植版またはそれに触発されたものです：
-
-### [9Router](https://github.com/decolua/9router)
-
-CLIProxyAPIに触発されたNext.js実装。インストールと使用が簡単で、フォーマット変換（OpenAI/Claude/Gemini/Ollama）、自動フォールバック付きコンボシステム、指数バックオフ付きマルチアカウント管理、Next.js Webダッシュボード、CLIツール（Cursor、Claude Code、Cline、RooCode）のサポートをゼロから構築 - APIキー不要
-
-### [OmniRoute](https://github.com/diegosouzapw/OmniRoute)
-
-コーディングを止めない。無料および低コストのAIモデルへのスマートルーティングと自動フォールバック。
-
-OmniRouteはマルチプロバイダーLLM向けのAIゲートウェイです：スマートルーティング、負荷分散、リトライ、フォールバックを備えたOpenAI互換エンドポイント。ポリシー、レート制限、キャッシュ、可観測性を追加して、信頼性が高くコストを意識した推論を実現します。
-
-> [!NOTE]
-> CLIProxyAPIの移植版またはそれに触発されたプロジェクトを開発した場合は、PRを送ってこのリストに追加してください。
-
-## ライセンス
-
-本プロジェクトはMITライセンスの下でライセンスされています - 詳細は[LICENSE](LICENSE)ファイルを参照してください。

config.example.yaml

@@ -96,6 +96,7 @@ max-retry-interval: 30
 quota-exceeded:
   switch-project: true # Whether to automatically switch to another project when a quota is exceeded
   switch-preview-model: true # Whether to automatically switch to a preview model when a quota is exceeded
+  antigravity-credits: true # Whether to retry Antigravity quota_exhausted 429s once with enabledCreditTypes=["GOOGLE_ONE_AI"]
 
 # Routing strategy for selecting credentials when multiple match.
 routing:
@@ -104,6 +105,10 @@ routing:
 # When true, enable authentication for the WebSocket API (/v1/ws).
 ws-auth: false
 
+# When true, enable Gemini CLI internal endpoints (/v1internal:*).
+# Default is false for safety.
+enable-gemini-cli-endpoint: false
+
 # When > 0, emit blank lines every N seconds for non-streaming responses to prevent idle timeouts.
 nonstream-keepalive-interval: 0
 
@@ -177,6 +182,8 @@ nonstream-keepalive-interval: 0
 #         - "API"
 #         - "proxy"
 #       cache-user-id: true          # optional: default is false; set true to reuse cached user_id per API key instead of generating a random one each request
+#     experimental-cch-signing: false # optional: default is false; when true, sign the final /v1/messages body using the current Claude Code cch algorithm
+#                                     # keep this disabled unless you explicitly need the behavior, so upstream seed changes fall back to legacy proxy behavior
 
 # Default headers for Claude API requests. Update when Claude Code releases new versions.
 # In legacy mode, user-agent/package-version/runtime-version/timeout are used as fallbacks
@@ -313,6 +320,10 @@ nonstream-keepalive-interval: 0
 # These aliases rename model IDs for both model listing and request routing.
 # Supported channels: gemini-cli, vertex, aistudio, antigravity, claude, codex, qwen, iflow, kiro, github-copilot, kimi.
 # NOTE: Aliases do not apply to gemini-api-key, codex-api-key, claude-api-key, openai-compatibility, vertex-api-key, or ampcode.
+# NOTE: Because aliases affect the merged /v1 model list and merged request routing, overlapping
+# client-visible names can become ambiguous across providers. /api/provider/{provider}/... helps
+# you select the protocol surface, but inference backend selection can still follow the resolved
+# model/alias. For strict backend pinning, use unique aliases/prefixes or avoid overlapping names.
 # You can repeat the same name with different aliases to expose multiple client model names.
 # oauth-model-alias:
 #  antigravity:

go.mod

@@ -83,6 +83,7 @@ require (
 	github.com/muesli/cancelreader v0.2.2 // indirect
 	github.com/muesli/termenv v0.16.0 // indirect
 	github.com/pelletier/go-toml/v2 v2.2.2 // indirect
+	github.com/pierrec/xxHash v0.1.5
 	github.com/pjbgf/sha1cd v0.5.0 // indirect
 	github.com/rivo/uniseg v0.4.7 // indirect
 	github.com/rs/xid v1.5.0 // indirect

go.sum

@@ -154,6 +154,8 @@ github.com/muesli/termenv v0.16.0 h1:S5AlUN9dENB57rsbnkPyfdGuWIlkmzJjbFf0Tf5FWUc
 github.com/muesli/termenv v0.16.0/go.mod h1:ZRfOIKPFDYQoDFF4Olj7/QJbW60Ol/kL1pU3VfY/Cnk=
 github.com/pelletier/go-toml/v2 v2.2.2 h1:aYUidT7k73Pcl9nb2gScu7NSrKCSHIDE89b3+6Wq+LM=
 github.com/pelletier/go-toml/v2 v2.2.2/go.mod h1:1t835xjRzz80PqgE6HHgN2JOsmgYu/h4qDAS4n929Rs=
+github.com/pierrec/xxHash v0.1.5 h1:n/jBpwTHiER4xYvK3/CdPVnLDPchj8eTJFFLUb4QHBo=
+github.com/pierrec/xxHash v0.1.5/go.mod h1:w2waW5Zoa/Wc4Yqe0wgrIYAGKqRMf7czn2HNKXmuL+I=
 github.com/pjbgf/sha1cd v0.5.0 h1:a+UkboSi1znleCDUNT3M5YxjOnN1fz2FhN48FlwCxs0=
 github.com/pjbgf/sha1cd v0.5.0/go.mod h1:lhpGlyHLpQZoxMv8HcgXvZEhcGs0PG/vsZnEJ7H0iCM=
 github.com/pkg/browser v0.0.0-20240102092130-5ac0b6a4141c h1:+mdjkGKdHQG3305AYmdv1U2eRNDiU2ErMBj1gwrq8eQ=

internal/api/handlers/management/auth_files.go

@@ -1047,6 +1047,7 @@ func (h *Handler) buildAuthFromFileData(path string, data []byte) (*coreauth.Aut
 			auth.Runtime = existing.Runtime
 		}
 	}
+	coreauth.ApplyCustomHeadersFromMetadata(auth)
 	return auth, nil
 }
 
@@ -1129,7 +1130,7 @@ func (h *Handler) PatchAuthFileStatus(c *gin.Context) {
 	c.JSON(http.StatusOK, gin.H{"status": "ok", "disabled": *req.Disabled})
 }
 
-// PatchAuthFileFields updates editable fields (prefix, proxy_url, priority, note) of an auth file.
+// PatchAuthFileFields updates editable fields (prefix, proxy_url, headers, priority, note) of an auth file.
 func (h *Handler) PatchAuthFileFields(c *gin.Context) {
 	if h.authManager == nil {
 		c.JSON(http.StatusServiceUnavailable, gin.H{"error": "core auth manager unavailable"})
@@ -1137,11 +1138,12 @@ func (h *Handler) PatchAuthFileFields(c *gin.Context) {
 	}
 
 	var req struct {
-		Name     string  `json:"name"`
-		Prefix   *string `json:"prefix"`
-		ProxyURL *string `json:"proxy_url"`
-		Priority *int    `json:"priority"`
-		Note     *string `json:"note"`
+		Name     string            `json:"name"`
+		Prefix   *string           `json:"prefix"`
+		ProxyURL *string           `json:"proxy_url"`
+		Headers  map[string]string `json:"headers"`
+		Priority *int              `json:"priority"`
+		Note     *string           `json:"note"`
 	}
 	if err := c.ShouldBindJSON(&req); err != nil {
 		c.JSON(http.StatusBadRequest, gin.H{"error": "invalid request body"})
@@ -1177,13 +1179,107 @@ func (h *Handler) PatchAuthFileFields(c *gin.Context) {
 
 	changed := false
 	if req.Prefix != nil {
-		targetAuth.Prefix = *req.Prefix
+		prefix := strings.TrimSpace(*req.Prefix)
+		targetAuth.Prefix = prefix
+		if targetAuth.Metadata == nil {
+			targetAuth.Metadata = make(map[string]any)
+		}
+		if prefix == "" {
+			delete(targetAuth.Metadata, "prefix")
+		} else {
+			targetAuth.Metadata["prefix"] = prefix
+		}
 		changed = true
 	}
 	if req.ProxyURL != nil {
-		targetAuth.ProxyURL = *req.ProxyURL
+		proxyURL := strings.TrimSpace(*req.ProxyURL)
+		targetAuth.ProxyURL = proxyURL
+		if targetAuth.Metadata == nil {
+			targetAuth.Metadata = make(map[string]any)
+		}
+		if proxyURL == "" {
+			delete(targetAuth.Metadata, "proxy_url")
+		} else {
+			targetAuth.Metadata["proxy_url"] = proxyURL
+		}
 		changed = true
 	}
+	if len(req.Headers) > 0 {
+		existingHeaders := coreauth.ExtractCustomHeadersFromMetadata(targetAuth.Metadata)
+		nextHeaders := make(map[string]string, len(existingHeaders))
+		for k, v := range existingHeaders {
+			nextHeaders[k] = v
+		}
+		headerChanged := false
+
+		for key, value := range req.Headers {
+			name := strings.TrimSpace(key)
+			if name == "" {
+				continue
+			}
+			val := strings.TrimSpace(value)
+			attrKey := "header:" + name
+			if val == "" {
+				if _, ok := nextHeaders[name]; ok {
+					delete(nextHeaders, name)
+					headerChanged = true
+				}
+				if targetAuth.Attributes != nil {
+					if _, ok := targetAuth.Attributes[attrKey]; ok {
+						headerChanged = true
+					}
+				}
+				continue
+			}
+			if prev, ok := nextHeaders[name]; !ok || prev != val {
+				headerChanged = true
+			}
+			nextHeaders[name] = val
+			if targetAuth.Attributes != nil {
+				if prev, ok := targetAuth.Attributes[attrKey]; !ok || prev != val {
+					headerChanged = true
+				}
+			} else {
+				headerChanged = true
+			}
+		}
+
+		if headerChanged {
+			if targetAuth.Metadata == nil {
+				targetAuth.Metadata = make(map[string]any)
+			}
+			if targetAuth.Attributes == nil {
+				targetAuth.Attributes = make(map[string]string)
+			}
+
+			for key, value := range req.Headers {
+				name := strings.TrimSpace(key)
+				if name == "" {
+					continue
+				}
+				val := strings.TrimSpace(value)
+				attrKey := "header:" + name
+				if val == "" {
+					delete(nextHeaders, name)
+					delete(targetAuth.Attributes, attrKey)
+					continue
+				}
+				nextHeaders[name] = val
+				targetAuth.Attributes[attrKey] = val
+			}
+
+			if len(nextHeaders) == 0 {
+				delete(targetAuth.Metadata, "headers")
+			} else {
+				metaHeaders := make(map[string]any, len(nextHeaders))
+				for k, v := range nextHeaders {
+					metaHeaders[k] = v
+				}
+				targetAuth.Metadata["headers"] = metaHeaders
+			}
+			changed = true
+		}
+	}
 	if req.Priority != nil || req.Note != nil {
 		if targetAuth.Metadata == nil {
 			targetAuth.Metadata = make(map[string]any)

internal/api/handlers/management/auth_files_patch_fields_test.go

@@ -0,0 +1,164 @@
+package management
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"github.com/gin-gonic/gin"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	coreauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+)
+
+func TestPatchAuthFileFields_MergeHeadersAndDeleteEmptyValues(t *testing.T) {
+	t.Setenv("MANAGEMENT_PASSWORD", "")
+	gin.SetMode(gin.TestMode)
+
+	store := &memoryAuthStore{}
+	manager := coreauth.NewManager(store, nil, nil)
+	record := &coreauth.Auth{
+		ID:       "test.json",
+		FileName: "test.json",
+		Provider: "claude",
+		Attributes: map[string]string{
+			"path":            "/tmp/test.json",
+			"header:X-Old":    "old",
+			"header:X-Remove": "gone",
+		},
+		Metadata: map[string]any{
+			"type": "claude",
+			"headers": map[string]any{
+				"X-Old":    "old",
+				"X-Remove": "gone",
+			},
+		},
+	}
+	if _, errRegister := manager.Register(context.Background(), record); errRegister != nil {
+		t.Fatalf("failed to register auth record: %v", errRegister)
+	}
+
+	h := NewHandlerWithoutConfigFilePath(&config.Config{AuthDir: t.TempDir()}, manager)
+
+	body := `{"name":"test.json","prefix":"p1","proxy_url":"http://proxy.local","headers":{"X-Old":"new","X-New":"v","X-Remove":"  ","X-Nope":""}}`
+	rec := httptest.NewRecorder()
+	ctx, _ := gin.CreateTestContext(rec)
+	req := httptest.NewRequest(http.MethodPatch, "/v0/management/auth-files/fields", strings.NewReader(body))
+	req.Header.Set("Content-Type", "application/json")
+	ctx.Request = req
+	h.PatchAuthFileFields(ctx)
+
+	if rec.Code != http.StatusOK {
+		t.Fatalf("expected status %d, got %d with body %s", http.StatusOK, rec.Code, rec.Body.String())
+	}
+
+	updated, ok := manager.GetByID("test.json")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth record to exist after patch")
+	}
+
+	if updated.Prefix != "p1" {
+		t.Fatalf("prefix = %q, want %q", updated.Prefix, "p1")
+	}
+	if updated.ProxyURL != "http://proxy.local" {
+		t.Fatalf("proxy_url = %q, want %q", updated.ProxyURL, "http://proxy.local")
+	}
+
+	if updated.Metadata == nil {
+		t.Fatalf("expected metadata to be non-nil")
+	}
+	if got, _ := updated.Metadata["prefix"].(string); got != "p1" {
+		t.Fatalf("metadata.prefix = %q, want %q", got, "p1")
+	}
+	if got, _ := updated.Metadata["proxy_url"].(string); got != "http://proxy.local" {
+		t.Fatalf("metadata.proxy_url = %q, want %q", got, "http://proxy.local")
+	}
+
+	headersMeta, ok := updated.Metadata["headers"].(map[string]any)
+	if !ok {
+		raw, _ := json.Marshal(updated.Metadata["headers"])
+		t.Fatalf("metadata.headers = %T (%s), want map[string]any", updated.Metadata["headers"], string(raw))
+	}
+	if got := headersMeta["X-Old"]; got != "new" {
+		t.Fatalf("metadata.headers.X-Old = %#v, want %q", got, "new")
+	}
+	if got := headersMeta["X-New"]; got != "v" {
+		t.Fatalf("metadata.headers.X-New = %#v, want %q", got, "v")
+	}
+	if _, ok := headersMeta["X-Remove"]; ok {
+		t.Fatalf("expected metadata.headers.X-Remove to be deleted")
+	}
+	if _, ok := headersMeta["X-Nope"]; ok {
+		t.Fatalf("expected metadata.headers.X-Nope to be absent")
+	}
+
+	if got := updated.Attributes["header:X-Old"]; got != "new" {
+		t.Fatalf("attrs header:X-Old = %q, want %q", got, "new")
+	}
+	if got := updated.Attributes["header:X-New"]; got != "v" {
+		t.Fatalf("attrs header:X-New = %q, want %q", got, "v")
+	}
+	if _, ok := updated.Attributes["header:X-Remove"]; ok {
+		t.Fatalf("expected attrs header:X-Remove to be deleted")
+	}
+	if _, ok := updated.Attributes["header:X-Nope"]; ok {
+		t.Fatalf("expected attrs header:X-Nope to be absent")
+	}
+}
+
+func TestPatchAuthFileFields_HeadersEmptyMapIsNoop(t *testing.T) {
+	t.Setenv("MANAGEMENT_PASSWORD", "")
+	gin.SetMode(gin.TestMode)
+
+	store := &memoryAuthStore{}
+	manager := coreauth.NewManager(store, nil, nil)
+	record := &coreauth.Auth{
+		ID:       "noop.json",
+		FileName: "noop.json",
+		Provider: "claude",
+		Attributes: map[string]string{
+			"path":         "/tmp/noop.json",
+			"header:X-Kee": "1",
+		},
+		Metadata: map[string]any{
+			"type": "claude",
+			"headers": map[string]any{
+				"X-Kee": "1",
+			},
+		},
+	}
+	if _, errRegister := manager.Register(context.Background(), record); errRegister != nil {
+		t.Fatalf("failed to register auth record: %v", errRegister)
+	}
+
+	h := NewHandlerWithoutConfigFilePath(&config.Config{AuthDir: t.TempDir()}, manager)
+
+	body := `{"name":"noop.json","note":"hello","headers":{}}`
+	rec := httptest.NewRecorder()
+	ctx, _ := gin.CreateTestContext(rec)
+	req := httptest.NewRequest(http.MethodPatch, "/v0/management/auth-files/fields", strings.NewReader(body))
+	req.Header.Set("Content-Type", "application/json")
+	ctx.Request = req
+	h.PatchAuthFileFields(ctx)
+
+	if rec.Code != http.StatusOK {
+		t.Fatalf("expected status %d, got %d with body %s", http.StatusOK, rec.Code, rec.Body.String())
+	}
+
+	updated, ok := manager.GetByID("noop.json")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth record to exist after patch")
+	}
+	if got := updated.Attributes["header:X-Kee"]; got != "1" {
+		t.Fatalf("attrs header:X-Kee = %q, want %q", got, "1")
+	}
+	headersMeta, ok := updated.Metadata["headers"].(map[string]any)
+	if !ok {
+		t.Fatalf("expected metadata.headers to remain a map, got %T", updated.Metadata["headers"])
+	}
+	if got := headersMeta["X-Kee"]; got != "1" {
+		t.Fatalf("metadata.headers.X-Kee = %#v, want %q", got, "1")
+	}
+}

internal/api/middleware/response_writer.go

@@ -15,6 +15,8 @@ import (
 )
 
 const requestBodyOverrideContextKey = "REQUEST_BODY_OVERRIDE"
+const responseBodyOverrideContextKey = "RESPONSE_BODY_OVERRIDE"
+const websocketTimelineOverrideContextKey = "WEBSOCKET_TIMELINE_OVERRIDE"
 
 // RequestInfo holds essential details of an incoming HTTP request for logging purposes.
 type RequestInfo struct {
@@ -304,6 +306,10 @@ func (w *ResponseWriterWrapper) Finalize(c *gin.Context) error {
 		if len(apiResponse) > 0 {
 			_ = w.streamWriter.WriteAPIResponse(apiResponse)
 		}
+		apiWebsocketTimeline := w.extractAPIWebsocketTimeline(c)
+		if len(apiWebsocketTimeline) > 0 {
+			_ = w.streamWriter.WriteAPIWebsocketTimeline(apiWebsocketTimeline)
+		}
 		if err := w.streamWriter.Close(); err != nil {
 			w.streamWriter = nil
 			return err
@@ -312,7 +318,7 @@ func (w *ResponseWriterWrapper) Finalize(c *gin.Context) error {
 		return nil
 	}
 
-	return w.logRequest(w.extractRequestBody(c), finalStatusCode, w.cloneHeaders(), w.body.Bytes(), w.extractAPIRequest(c), w.extractAPIResponse(c), w.extractAPIResponseTimestamp(c), slicesAPIResponseError, forceLog)
+	return w.logRequest(w.extractRequestBody(c), finalStatusCode, w.cloneHeaders(), w.extractResponseBody(c), w.extractWebsocketTimeline(c), w.extractAPIRequest(c), w.extractAPIResponse(c), w.extractAPIWebsocketTimeline(c), w.extractAPIResponseTimestamp(c), slicesAPIResponseError, forceLog)
 }
 
 func (w *ResponseWriterWrapper) cloneHeaders() map[string][]string {
@@ -352,6 +358,18 @@ func (w *ResponseWriterWrapper) extractAPIResponse(c *gin.Context) []byte {
 	return data
 }
 
+func (w *ResponseWriterWrapper) extractAPIWebsocketTimeline(c *gin.Context) []byte {
+	apiTimeline, isExist := c.Get("API_WEBSOCKET_TIMELINE")
+	if !isExist {
+		return nil
+	}
+	data, ok := apiTimeline.([]byte)
+	if !ok || len(data) == 0 {
+		return nil
+	}
+	return bytes.Clone(data)
+}
+
 func (w *ResponseWriterWrapper) extractAPIResponseTimestamp(c *gin.Context) time.Time {
 	ts, isExist := c.Get("API_RESPONSE_TIMESTAMP")
 	if !isExist {
@@ -364,19 +382,8 @@ func (w *ResponseWriterWrapper) extractAPIResponseTimestamp(c *gin.Context) time
 }
 
 func (w *ResponseWriterWrapper) extractRequestBody(c *gin.Context) []byte {
-	if c != nil {
-		if bodyOverride, isExist := c.Get(requestBodyOverrideContextKey); isExist {
-			switch value := bodyOverride.(type) {
-			case []byte:
-				if len(value) > 0 {
-					return bytes.Clone(value)
-				}
-			case string:
-				if strings.TrimSpace(value) != "" {
-					return []byte(value)
-				}
-			}
-		}
+	if body := extractBodyOverride(c, requestBodyOverrideContextKey); len(body) > 0 {
+		return body
 	}
 	if w.requestInfo != nil && len(w.requestInfo.Body) > 0 {
 		return w.requestInfo.Body
@@ -384,13 +391,48 @@ func (w *ResponseWriterWrapper) extractRequestBody(c *gin.Context) []byte {
 	return nil
 }
 
-func (w *ResponseWriterWrapper) logRequest(requestBody []byte, statusCode int, headers map[string][]string, body []byte, apiRequestBody, apiResponseBody []byte, apiResponseTimestamp time.Time, apiResponseErrors []*interfaces.ErrorMessage, forceLog bool) error {
+func (w *ResponseWriterWrapper) extractResponseBody(c *gin.Context) []byte {
+	if body := extractBodyOverride(c, responseBodyOverrideContextKey); len(body) > 0 {
+		return body
+	}
+	if w.body == nil || w.body.Len() == 0 {
+		return nil
+	}
+	return bytes.Clone(w.body.Bytes())
+}
+
+func (w *ResponseWriterWrapper) extractWebsocketTimeline(c *gin.Context) []byte {
+	return extractBodyOverride(c, websocketTimelineOverrideContextKey)
+}
+
+func extractBodyOverride(c *gin.Context, key string) []byte {
+	if c == nil {
+		return nil
+	}
+	bodyOverride, isExist := c.Get(key)
+	if !isExist {
+		return nil
+	}
+	switch value := bodyOverride.(type) {
+	case []byte:
+		if len(value) > 0 {
+			return bytes.Clone(value)
+		}
+	case string:
+		if strings.TrimSpace(value) != "" {
+			return []byte(value)
+		}
+	}
+	return nil
+}
+
+func (w *ResponseWriterWrapper) logRequest(requestBody []byte, statusCode int, headers map[string][]string, body, websocketTimeline, apiRequestBody, apiResponseBody, apiWebsocketTimeline []byte, apiResponseTimestamp time.Time, apiResponseErrors []*interfaces.ErrorMessage, forceLog bool) error {
 	if w.requestInfo == nil {
 		return nil
 	}
 
 	if loggerWithOptions, ok := w.logger.(interface {
-		LogRequestWithOptions(string, string, map[string][]string, []byte, int, map[string][]string, []byte, []byte, []byte, []*interfaces.ErrorMessage, bool, string, time.Time, time.Time) error
+		LogRequestWithOptions(string, string, map[string][]string, []byte, int, map[string][]string, []byte, []byte, []byte, []byte, []byte, []*interfaces.ErrorMessage, bool, string, time.Time, time.Time) error
 	}); ok {
 		return loggerWithOptions.LogRequestWithOptions(
 			w.requestInfo.URL,
@@ -400,8 +442,10 @@ func (w *ResponseWriterWrapper) logRequest(requestBody []byte, statusCode int, h
 			statusCode,
 			headers,
 			body,
+			websocketTimeline,
 			apiRequestBody,
 			apiResponseBody,
+			apiWebsocketTimeline,
 			apiResponseErrors,
 			forceLog,
 			w.requestInfo.RequestID,
@@ -418,8 +462,10 @@ func (w *ResponseWriterWrapper) logRequest(requestBody []byte, statusCode int, h
 		statusCode,
 		headers,
 		body,
+		websocketTimeline,
 		apiRequestBody,
 		apiResponseBody,
+		apiWebsocketTimeline,
 		apiResponseErrors,
 		w.requestInfo.RequestID,
 		w.requestInfo.Timestamp,

internal/api/middleware/response_writer_test.go

@@ -1,10 +1,14 @@
 package middleware
 
 import (
+	"bytes"
 	"net/http/httptest"
 	"testing"
+	"time"
 
 	"github.com/gin-gonic/gin"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/interfaces"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/logging"
 )
 
 func TestExtractRequestBodyPrefersOverride(t *testing.T) {
@@ -33,7 +37,7 @@ func TestExtractRequestBodySupportsStringOverride(t *testing.T) {
 	recorder := httptest.NewRecorder()
 	c, _ := gin.CreateTestContext(recorder)
 
-	wrapper := &ResponseWriterWrapper{}
+	wrapper := &ResponseWriterWrapper{body: &bytes.Buffer{}}
 	c.Set(requestBodyOverrideContextKey, "override-as-string")
 
 	body := wrapper.extractRequestBody(c)
@@ -41,3 +45,158 @@ func TestExtractRequestBodySupportsStringOverride(t *testing.T) {
 		t.Fatalf("request body = %q, want %q", string(body), "override-as-string")
 	}
 }
+
+func TestExtractResponseBodyPrefersOverride(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+
+	wrapper := &ResponseWriterWrapper{body: &bytes.Buffer{}}
+	wrapper.body.WriteString("original-response")
+
+	body := wrapper.extractResponseBody(c)
+	if string(body) != "original-response" {
+		t.Fatalf("response body = %q, want %q", string(body), "original-response")
+	}
+
+	c.Set(responseBodyOverrideContextKey, []byte("override-response"))
+	body = wrapper.extractResponseBody(c)
+	if string(body) != "override-response" {
+		t.Fatalf("response body = %q, want %q", string(body), "override-response")
+	}
+
+	body[0] = 'X'
+	if got := wrapper.extractResponseBody(c); string(got) != "override-response" {
+		t.Fatalf("response override should be cloned, got %q", string(got))
+	}
+}
+
+func TestExtractResponseBodySupportsStringOverride(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+
+	wrapper := &ResponseWriterWrapper{}
+	c.Set(responseBodyOverrideContextKey, "override-response-as-string")
+
+	body := wrapper.extractResponseBody(c)
+	if string(body) != "override-response-as-string" {
+		t.Fatalf("response body = %q, want %q", string(body), "override-response-as-string")
+	}
+}
+
+func TestExtractBodyOverrideClonesBytes(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+
+	override := []byte("body-override")
+	c.Set(requestBodyOverrideContextKey, override)
+
+	body := extractBodyOverride(c, requestBodyOverrideContextKey)
+	if !bytes.Equal(body, override) {
+		t.Fatalf("body override = %q, want %q", string(body), string(override))
+	}
+
+	body[0] = 'X'
+	if !bytes.Equal(override, []byte("body-override")) {
+		t.Fatalf("override mutated: %q", string(override))
+	}
+}
+
+func TestExtractWebsocketTimelineUsesOverride(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+
+	wrapper := &ResponseWriterWrapper{}
+	if got := wrapper.extractWebsocketTimeline(c); got != nil {
+		t.Fatalf("expected nil websocket timeline, got %q", string(got))
+	}
+
+	c.Set(websocketTimelineOverrideContextKey, []byte("timeline"))
+	body := wrapper.extractWebsocketTimeline(c)
+	if string(body) != "timeline" {
+		t.Fatalf("websocket timeline = %q, want %q", string(body), "timeline")
+	}
+}
+
+func TestFinalizeStreamingWritesAPIWebsocketTimeline(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+
+	streamWriter := &testStreamingLogWriter{}
+	wrapper := &ResponseWriterWrapper{
+		ResponseWriter: c.Writer,
+		logger:         &testRequestLogger{enabled: true},
+		requestInfo: &RequestInfo{
+			URL:       "/v1/responses",
+			Method:    "POST",
+			Headers:   map[string][]string{"Content-Type": {"application/json"}},
+			RequestID: "req-1",
+			Timestamp: time.Date(2026, time.April, 1, 12, 0, 0, 0, time.UTC),
+		},
+		isStreaming:  true,
+		streamWriter: streamWriter,
+	}
+
+	c.Set("API_WEBSOCKET_TIMELINE", []byte("Timestamp: 2026-04-01T12:00:00Z\nEvent: api.websocket.request\n{}"))
+
+	if err := wrapper.Finalize(c); err != nil {
+		t.Fatalf("Finalize error: %v", err)
+	}
+	if string(streamWriter.apiWebsocketTimeline) != "Timestamp: 2026-04-01T12:00:00Z\nEvent: api.websocket.request\n{}" {
+		t.Fatalf("stream writer websocket timeline = %q", string(streamWriter.apiWebsocketTimeline))
+	}
+	if !streamWriter.closed {
+		t.Fatal("expected stream writer to be closed")
+	}
+}
+
+type testRequestLogger struct {
+	enabled bool
+}
+
+func (l *testRequestLogger) LogRequest(string, string, map[string][]string, []byte, int, map[string][]string, []byte, []byte, []byte, []byte, []byte, []*interfaces.ErrorMessage, string, time.Time, time.Time) error {
+	return nil
+}
+
+func (l *testRequestLogger) LogStreamingRequest(string, string, map[string][]string, []byte, string) (logging.StreamingLogWriter, error) {
+	return &testStreamingLogWriter{}, nil
+}
+
+func (l *testRequestLogger) IsEnabled() bool {
+	return l.enabled
+}
+
+type testStreamingLogWriter struct {
+	apiWebsocketTimeline []byte
+	closed               bool
+}
+
+func (w *testStreamingLogWriter) WriteChunkAsync([]byte) {}
+
+func (w *testStreamingLogWriter) WriteStatus(int, map[string][]string) error {
+	return nil
+}
+
+func (w *testStreamingLogWriter) WriteAPIRequest([]byte) error {
+	return nil
+}
+
+func (w *testStreamingLogWriter) WriteAPIResponse([]byte) error {
+	return nil
+}
+
+func (w *testStreamingLogWriter) WriteAPIWebsocketTimeline(apiWebsocketTimeline []byte) error {
+	w.apiWebsocketTimeline = bytes.Clone(apiWebsocketTimeline)
+	return nil
+}
+
+func (w *testStreamingLogWriter) SetFirstChunkTimestamp(time.Time) {}
+
+func (w *testStreamingLogWriter) Close() error {
+	w.closed = true
+	return nil
+}

internal/api/modules/amp/fallback_handlers.go

@@ -123,6 +123,10 @@ func (fh *FallbackHandler) WrapHandler(handler gin.HandlerFunc) gin.HandlerFunc
 			return
 		}
 
+		// Sanitize request body: remove thinking blocks with invalid signatures
+		// to prevent upstream API 400 errors
+		bodyBytes = SanitizeAmpRequestBody(bodyBytes)
+
 		// Restore the body for the handler to read
 		c.Request.Body = io.NopCloser(bytes.NewReader(bodyBytes))
 
@@ -249,6 +253,7 @@ func (fh *FallbackHandler) WrapHandler(handler gin.HandlerFunc) gin.HandlerFunc
 			log.Debugf("amp model mapping: request %s -> %s", normalizedModel, resolvedModel)
 			logAmpRouting(RouteTypeModelMapping, modelName, resolvedModel, providerName, requestPath)
 			rewriter := NewResponseRewriter(c.Writer, modelName)
+			rewriter.suppressThinking = true
 			c.Writer = rewriter
 			// Filter Anthropic-Beta header only for local handling paths
 			filterAntropicBetaHeader(c)
@@ -259,10 +264,17 @@ func (fh *FallbackHandler) WrapHandler(handler gin.HandlerFunc) gin.HandlerFunc
 		} else if len(providers) > 0 {
 			// Log: Using local provider (free)
 			logAmpRouting(RouteTypeLocalProvider, modelName, resolvedModel, providerName, requestPath)
+			// Wrap with ResponseRewriter for local providers too, because upstream
+			// proxies (e.g. NewAPI) may return a different model name and lack
+			// Amp-required fields like thinking.signature.
+			rewriter := NewResponseRewriter(c.Writer, modelName)
+			rewriter.suppressThinking = providerName != "claude"
+			c.Writer = rewriter
 			// Filter Anthropic-Beta header only for local handling paths
 			filterAntropicBetaHeader(c)
 			c.Request.Body = io.NopCloser(bytes.NewReader(bodyBytes))
 			handler(c)
+			rewriter.Flush()
 		} else {
 			// No provider, no mapping, no proxy: fall back to the wrapped handler so it can return an error response
 			c.Request.Body = io.NopCloser(bytes.NewReader(bodyBytes))

internal/api/modules/amp/response_rewriter.go

@@ -2,6 +2,7 @@ package amp
 
 import (
 	"bytes"
+	"fmt"
 	"net/http"
 	"strings"
 
@@ -12,15 +13,17 @@ import (
 )
 
 // ResponseRewriter wraps a gin.ResponseWriter to intercept and modify the response body
-// It's used to rewrite model names in responses when model mapping is used
+// It is used to rewrite model names in responses when model mapping is used
+// and to keep Amp-compatible response shapes.
 type ResponseRewriter struct {
 	gin.ResponseWriter
-	body          *bytes.Buffer
-	originalModel string
-	isStreaming   bool
+	body             *bytes.Buffer
+	originalModel    string
+	isStreaming      bool
+	suppressThinking bool
 }
 
-// NewResponseRewriter creates a new response rewriter for model name substitution
+// NewResponseRewriter creates a new response rewriter for model name substitution.
 func NewResponseRewriter(w gin.ResponseWriter, originalModel string) *ResponseRewriter {
 	return &ResponseRewriter{
 		ResponseWriter: w,
@@ -33,15 +36,15 @@ const maxBufferedResponseBytes = 2 * 1024 * 1024 // 2MB safety cap
 
 func looksLikeSSEChunk(data []byte) bool {
 	// Fallback detection: some upstreams may omit/lie about Content-Type, causing SSE to be buffered.
-	// Heuristics are intentionally simple and cheap.
-	return bytes.Contains(data, []byte("data:")) ||
-		bytes.Contains(data, []byte("event:")) ||
-		bytes.Contains(data, []byte("message_start")) ||
-		bytes.Contains(data, []byte("message_delta")) ||
-		bytes.Contains(data, []byte("content_block_start")) ||
-		bytes.Contains(data, []byte("content_block_delta")) ||
-		bytes.Contains(data, []byte("content_block_stop")) ||
-		bytes.Contains(data, []byte("\n\n"))
+	// We conservatively detect SSE by checking for "data:" / "event:" at the start of any line.
+	for _, line := range bytes.Split(data, []byte("\n")) {
+		trimmed := bytes.TrimSpace(line)
+		if bytes.HasPrefix(trimmed, []byte("data:")) ||
+			bytes.HasPrefix(trimmed, []byte("event:")) {
+			return true
+		}
+	}
+	return false
 }
 
 func (rw *ResponseRewriter) enableStreaming(reason string) error {
@@ -95,7 +98,8 @@ func (rw *ResponseRewriter) Write(data []byte) (int, error) {
 	}
 
 	if rw.isStreaming {
-		n, err := rw.ResponseWriter.Write(rw.rewriteStreamChunk(data))
+		rewritten := rw.rewriteStreamChunk(data)
+		n, err := rw.ResponseWriter.Write(rewritten)
 		if err == nil {
 			if flusher, ok := rw.ResponseWriter.(http.Flusher); ok {
 				flusher.Flush()
@@ -106,7 +110,6 @@ func (rw *ResponseRewriter) Write(data []byte) (int, error) {
 	return rw.body.Write(data)
 }
 
-// Flush writes the buffered response with model names rewritten
 func (rw *ResponseRewriter) Flush() {
 	if rw.isStreaming {
 		if flusher, ok := rw.ResponseWriter.(http.Flusher); ok {
@@ -115,40 +118,79 @@ func (rw *ResponseRewriter) Flush() {
 		return
 	}
 	if rw.body.Len() > 0 {
-		if _, err := rw.ResponseWriter.Write(rw.rewriteModelInResponse(rw.body.Bytes())); err != nil {
+		rewritten := rw.rewriteModelInResponse(rw.body.Bytes())
+		// Update Content-Length to match the rewritten body size, since
+		// signature injection and model name changes alter the payload length.
+		rw.ResponseWriter.Header().Set("Content-Length", fmt.Sprintf("%d", len(rewritten)))
+		if _, err := rw.ResponseWriter.Write(rewritten); err != nil {
 			log.Warnf("amp response rewriter: failed to write rewritten response: %v", err)
 		}
 	}
 }
 
-// modelFieldPaths lists all JSON paths where model name may appear
 var modelFieldPaths = []string{"message.model", "model", "modelVersion", "response.model", "response.modelVersion"}
 
-// rewriteModelInResponse replaces all occurrences of the mapped model with the original model in JSON
-// It also suppresses "thinking" blocks if "tool_use" is present to ensure Amp client compatibility
-func (rw *ResponseRewriter) rewriteModelInResponse(data []byte) []byte {
-	// 1. Amp Compatibility: Suppress thinking blocks if tool use is detected
-	// The Amp client struggles when both thinking and tool_use blocks are present
+// ensureAmpSignature injects empty signature fields into tool_use/thinking blocks
+// in API responses so that the Amp TUI does not crash on P.signature.length.
+func ensureAmpSignature(data []byte) []byte {
+	for index, block := range gjson.GetBytes(data, "content").Array() {
+		blockType := block.Get("type").String()
+		if blockType != "tool_use" && blockType != "thinking" {
+			continue
+		}
+		signaturePath := fmt.Sprintf("content.%d.signature", index)
+		if gjson.GetBytes(data, signaturePath).Exists() {
+			continue
+		}
+		var err error
+		data, err = sjson.SetBytes(data, signaturePath, "")
+		if err != nil {
+			log.Warnf("Amp ResponseRewriter: failed to add empty signature to %s block: %v", blockType, err)
+			break
+		}
+	}
+
+	contentBlockType := gjson.GetBytes(data, "content_block.type").String()
+	if (contentBlockType == "tool_use" || contentBlockType == "thinking") && !gjson.GetBytes(data, "content_block.signature").Exists() {
+		var err error
+		data, err = sjson.SetBytes(data, "content_block.signature", "")
+		if err != nil {
+			log.Warnf("Amp ResponseRewriter: failed to add empty signature to streaming %s block: %v", contentBlockType, err)
+		}
+	}
+
+	return data
+}
+
+func (rw *ResponseRewriter) suppressAmpThinking(data []byte) []byte {
+	if !rw.suppressThinking {
+		return data
+	}
 	if gjson.GetBytes(data, `content.#(type=="tool_use")`).Exists() {
 		filtered := gjson.GetBytes(data, `content.#(type!="thinking")#`)
 		if filtered.Exists() {
 			originalCount := gjson.GetBytes(data, "content.#").Int()
 			filteredCount := filtered.Get("#").Int()
-
 			if originalCount > filteredCount {
 				var err error
 				data, err = sjson.SetBytes(data, "content", filtered.Value())
 				if err != nil {
 					log.Warnf("Amp ResponseRewriter: failed to suppress thinking blocks: %v", err)
-				} else {
-					log.Debugf("Amp ResponseRewriter: Suppressed %d thinking blocks due to tool usage", originalCount-filteredCount)
-					// Log the result for verification
-					log.Debugf("Amp ResponseRewriter: Resulting content: %s", gjson.GetBytes(data, "content").String())
 				}
 			}
 		}
 	}
 
+	return data
+}
+
+func (rw *ResponseRewriter) rewriteModelInResponse(data []byte) []byte {
+	data = ensureAmpSignature(data)
+	data = rw.suppressAmpThinking(data)
+	if len(data) == 0 {
+		return data
+	}
+
 	if rw.originalModel == "" {
 		return data
 	}
@@ -160,24 +202,154 @@ func (rw *ResponseRewriter) rewriteModelInResponse(data []byte) []byte {
 	return data
 }
 
-// rewriteStreamChunk rewrites model names in SSE stream chunks
 func (rw *ResponseRewriter) rewriteStreamChunk(chunk []byte) []byte {
-	if rw.originalModel == "" {
-		return chunk
+	lines := bytes.Split(chunk, []byte("\n"))
+	var out [][]byte
+
+	i := 0
+	for i < len(lines) {
+		line := lines[i]
+		trimmed := bytes.TrimSpace(line)
+
+		// Case 1: "event:" line - look ahead for its "data:" line
+		if bytes.HasPrefix(trimmed, []byte("event: ")) {
+			// Scan forward past blank lines to find the data: line
+			dataIdx := -1
+			for j := i + 1; j < len(lines); j++ {
+				t := bytes.TrimSpace(lines[j])
+				if len(t) == 0 {
+					continue
+				}
+				if bytes.HasPrefix(t, []byte("data: ")) {
+					dataIdx = j
+				}
+				break
+			}
+
+			if dataIdx >= 0 {
+				// Found event+data pair - process through rewriter
+				jsonData := bytes.TrimPrefix(bytes.TrimSpace(lines[dataIdx]), []byte("data: "))
+				if len(jsonData) > 0 && jsonData[0] == '{' {
+					rewritten := rw.rewriteStreamEvent(jsonData)
+					if rewritten == nil {
+						i = dataIdx + 1
+						continue
+					}
+					// Emit event line
+					out = append(out, line)
+					// Emit blank lines between event and data
+					for k := i + 1; k < dataIdx; k++ {
+						out = append(out, lines[k])
+					}
+					// Emit rewritten data
+					out = append(out, append([]byte("data: "), rewritten...))
+					i = dataIdx + 1
+					continue
+				}
+			}
+
+			// No data line found (orphan event from cross-chunk split)
+			// Pass it through as-is - the data will arrive in the next chunk
+			out = append(out, line)
+			i++
+			continue
+		}
+
+		// Case 2: standalone "data:" line (no preceding event: in this chunk)
+		if bytes.HasPrefix(trimmed, []byte("data: ")) {
+			jsonData := bytes.TrimPrefix(trimmed, []byte("data: "))
+			if len(jsonData) > 0 && jsonData[0] == '{' {
+				rewritten := rw.rewriteStreamEvent(jsonData)
+				if rewritten != nil {
+					out = append(out, append([]byte("data: "), rewritten...))
+				}
+				i++
+				continue
+			}
+		}
+
+		// Case 3: everything else
+		out = append(out, line)
+		i++
 	}
 
-	// SSE format: "data: {json}\n\n"
-	lines := bytes.Split(chunk, []byte("\n"))
-	for i, line := range lines {
-		if bytes.HasPrefix(line, []byte("data: ")) {
-			jsonData := bytes.TrimPrefix(line, []byte("data: "))
-			if len(jsonData) > 0 && jsonData[0] == '{' {
-				// Rewrite JSON in the data line
-				rewritten := rw.rewriteModelInResponse(jsonData)
-				lines[i] = append([]byte("data: "), rewritten...)
+	return bytes.Join(out, []byte("\n"))
+}
+
+// rewriteStreamEvent processes a single JSON event in the SSE stream.
+// It rewrites model names and ensures signature fields exist.
+// NOTE: streaming mode does NOT suppress thinking blocks - they are
+// passed through with signature injection to avoid breaking SSE index
+// alignment and TUI rendering.
+func (rw *ResponseRewriter) rewriteStreamEvent(data []byte) []byte {
+	// Inject empty signature where needed
+	data = ensureAmpSignature(data)
+
+	// Rewrite model name
+	if rw.originalModel != "" {
+		for _, path := range modelFieldPaths {
+			if gjson.GetBytes(data, path).Exists() {
+				data, _ = sjson.SetBytes(data, path, rw.originalModel)
 			}
 		}
 	}
 
-	return bytes.Join(lines, []byte("\n"))
+	return data
+}
+
+// SanitizeAmpRequestBody removes thinking blocks with empty/missing/invalid signatures
+// from the messages array in a request body before forwarding to the upstream API.
+// This prevents 400 errors from the API which requires valid signatures on thinking blocks.
+func SanitizeAmpRequestBody(body []byte) []byte {
+	messages := gjson.GetBytes(body, "messages")
+	if !messages.Exists() || !messages.IsArray() {
+		return body
+	}
+
+	modified := false
+	for msgIdx, msg := range messages.Array() {
+		if msg.Get("role").String() != "assistant" {
+			continue
+		}
+		content := msg.Get("content")
+		if !content.Exists() || !content.IsArray() {
+			continue
+		}
+
+		var keepBlocks []interface{}
+		removedCount := 0
+
+		for _, block := range content.Array() {
+			blockType := block.Get("type").String()
+			if blockType == "thinking" {
+				sig := block.Get("signature")
+				if !sig.Exists() || sig.Type != gjson.String || strings.TrimSpace(sig.String()) == "" {
+					removedCount++
+					continue
+				}
+			}
+			keepBlocks = append(keepBlocks, block.Value())
+		}
+
+		if removedCount > 0 {
+			contentPath := fmt.Sprintf("messages.%d.content", msgIdx)
+			var err error
+			if len(keepBlocks) == 0 {
+				body, err = sjson.SetBytes(body, contentPath, []interface{}{})
+			} else {
+				body, err = sjson.SetBytes(body, contentPath, keepBlocks)
+			}
+			if err != nil {
+				log.Warnf("Amp RequestSanitizer: failed to remove thinking blocks from message %d: %v", msgIdx, err)
+				continue
+			}
+			modified = true
+			log.Debugf("Amp RequestSanitizer: removed %d thinking blocks with invalid signatures from message %d", removedCount, msgIdx)
+		}
+	}
+
+	if modified {
+		log.Debugf("Amp RequestSanitizer: sanitized request body")
+	}
+	return body
 }

internal/api/modules/amp/response_rewriter_test.go

@@ -1,6 +1,7 @@
 package amp
 
 import (
+	"strings"
 	"testing"
 )
 
@@ -100,6 +101,50 @@ func TestRewriteStreamChunk_MessageModel(t *testing.T) {
 	}
 }
 
+func TestRewriteStreamChunk_PreservesThinkingWithSignatureInjection(t *testing.T) {
+	rw := &ResponseRewriter{}
+
+	chunk := []byte("event: content_block_start\ndata: {\"type\":\"content_block_start\",\"index\":0,\"content_block\":{\"type\":\"thinking\",\"thinking\":\"\"}}\n\nevent: content_block_delta\ndata: {\"type\":\"content_block_delta\",\"index\":0,\"delta\":{\"type\":\"thinking_delta\",\"thinking\":\"abc\"}}\n\nevent: content_block_stop\ndata: {\"type\":\"content_block_stop\",\"index\":0}\n\nevent: content_block_start\ndata: {\"type\":\"content_block_start\",\"index\":1,\"content_block\":{\"type\":\"tool_use\",\"name\":\"bash\",\"input\":{}}}\n\n")
+	result := rw.rewriteStreamChunk(chunk)
+
+	// Streaming mode preserves thinking blocks (does NOT suppress them)
+	// to avoid breaking SSE index alignment and TUI rendering
+	if !contains(result, []byte(`"content_block":{"type":"thinking"`)) {
+		t.Fatalf("expected thinking content_block_start to be preserved, got %s", string(result))
+	}
+	if !contains(result, []byte(`"delta":{"type":"thinking_delta"`)) {
+		t.Fatalf("expected thinking_delta to be preserved, got %s", string(result))
+	}
+	if !contains(result, []byte(`"type":"content_block_stop","index":0`)) {
+		t.Fatalf("expected content_block_stop for thinking block to be preserved, got %s", string(result))
+	}
+	if !contains(result, []byte(`"content_block":{"type":"tool_use"`)) {
+		t.Fatalf("expected tool_use content_block frame to remain, got %s", string(result))
+	}
+	// Signature should be injected into both thinking and tool_use blocks
+	if count := strings.Count(string(result), `"signature":""`); count != 2 {
+		t.Fatalf("expected 2 signature injections, but got %d in %s", count, string(result))
+	}
+}
+
+func TestSanitizeAmpRequestBody_RemovesWhitespaceAndNonStringSignatures(t *testing.T) {
+	input := []byte(`{"messages":[{"role":"assistant","content":[{"type":"thinking","thinking":"drop-whitespace","signature":"   "},{"type":"thinking","thinking":"drop-number","signature":123},{"type":"thinking","thinking":"keep-valid","signature":"valid-signature"},{"type":"text","text":"keep-text"}]}]}`)
+	result := SanitizeAmpRequestBody(input)
+
+	if contains(result, []byte("drop-whitespace")) {
+		t.Fatalf("expected whitespace-only signature block to be removed, got %s", string(result))
+	}
+	if contains(result, []byte("drop-number")) {
+		t.Fatalf("expected non-string signature block to be removed, got %s", string(result))
+	}
+	if !contains(result, []byte("keep-valid")) {
+		t.Fatalf("expected valid thinking block to remain, got %s", string(result))
+	}
+	if !contains(result, []byte("keep-text")) {
+		t.Fatalf("expected non-thinking content to remain, got %s", string(result))
+	}
+}
+
 func contains(data, substr []byte) bool {
 	for i := 0; i <= len(data)-len(substr); i++ {
 		if string(data[i:i+len(substr)]) == string(substr) {

internal/api/server.go

@@ -323,6 +323,10 @@ func NewServer(cfg *config.Config, authManager *auth.Manager, accessManager *sdk
 // setupRoutes configures the API routes for the server.
 // It defines the endpoints and associates them with their respective handlers.
 func (s *Server) setupRoutes() {
+	s.engine.GET("/healthz", func(c *gin.Context) {
+		c.JSON(http.StatusOK, gin.H{"status": "ok"})
+	})
+
 	s.engine.GET("/management.html", s.serveManagementControlPanel)
 	openaiHandlers := openai.NewOpenAIAPIHandler(s.handlers)
 	geminiHandlers := gemini.NewGeminiAPIHandler(s.handlers)
@@ -569,6 +573,8 @@ func (s *Server) registerManagementRoutes() {
 		mgmt.PUT("/quota-exceeded/switch-preview-model", s.mgmt.PutSwitchPreviewModel)
 		mgmt.PATCH("/quota-exceeded/switch-preview-model", s.mgmt.PutSwitchPreviewModel)
 
+		mgmt.GET("/copilot-quota", s.mgmt.GetCopilotQuota)
+
 		mgmt.GET("/api-keys", s.mgmt.GetAPIKeys)
 		mgmt.PUT("/api-keys", s.mgmt.PutAPIKeys)
 		mgmt.PATCH("/api-keys", s.mgmt.PatchAPIKeys)

internal/api/server_test.go

@@ -1,6 +1,7 @@
 package api
 
 import (
+	"encoding/json"
 	"net/http"
 	"net/http/httptest"
 	"os"
@@ -46,6 +47,28 @@ func newTestServer(t *testing.T) *Server {
 	return NewServer(cfg, authManager, accessManager, configPath)
 }
 
+func TestHealthz(t *testing.T) {
+	server := newTestServer(t)
+
+	req := httptest.NewRequest(http.MethodGet, "/healthz", nil)
+	rr := httptest.NewRecorder()
+	server.engine.ServeHTTP(rr, req)
+
+	if rr.Code != http.StatusOK {
+		t.Fatalf("unexpected status code: got %d want %d; body=%s", rr.Code, http.StatusOK, rr.Body.String())
+	}
+
+	var resp struct {
+		Status string `json:"status"`
+	}
+	if err := json.Unmarshal(rr.Body.Bytes(), &resp); err != nil {
+		t.Fatalf("failed to parse response JSON: %v; body=%s", err, rr.Body.String())
+	}
+	if resp.Status != "ok" {
+		t.Fatalf("unexpected response status: got %q want %q", resp.Status, "ok")
+	}
+}
+
 func TestAmpProviderModelRoutes(t *testing.T) {
 	testCases := []struct {
 		name         string
@@ -172,6 +195,8 @@ func TestDefaultRequestLoggerFactory_UsesResolvedLogDirectory(t *testing.T) {
 		nil,
 		nil,
 		nil,
+		nil,
+		nil,
 		true,
 		"issue-1711",
 		time.Now(),

internal/auth/claude/anthropic_auth.go

@@ -88,7 +88,7 @@ func (o *ClaudeAuth) GenerateAuthURL(state string, pkceCodes *PKCECodes) (string
 		"client_id":             {ClientID},
 		"response_type":         {"code"},
 		"redirect_uri":          {RedirectURI},
-		"scope":                 {"org:create_api_key user:profile user:inference"},
+		"scope":                 {"user:profile user:inference user:sessions:claude_code user:mcp_servers user:file_upload"},
 		"code_challenge":        {pkceCodes.CodeChallenge},
 		"code_challenge_method": {"S256"},
 		"state":                 {state},

internal/auth/copilot/copilot_auth.go

@@ -235,6 +235,74 @@ type CopilotModelEntry struct {
 	Capabilities map[string]any `json:"capabilities,omitempty"`
 }
 
+// CopilotModelLimits holds the token limits returned by the Copilot /models API
+// under capabilities.limits. These limits vary by account type (individual vs
+// business) and are the authoritative source for enforcing prompt size.
+type CopilotModelLimits struct {
+	// MaxContextWindowTokens is the total context window (prompt + output).
+	MaxContextWindowTokens int
+	// MaxPromptTokens is the hard limit on input/prompt tokens.
+	// Exceeding this triggers a 400 error from the Copilot API.
+	MaxPromptTokens int
+	// MaxOutputTokens is the maximum number of output/completion tokens.
+	MaxOutputTokens int
+}
+
+// Limits extracts the token limits from the model's capabilities map.
+// Returns nil if no limits are available or the structure is unexpected.
+//
+// Expected Copilot API shape:
+//
+//	"capabilities": {
+//	    "limits": {
+//	        "max_context_window_tokens": 200000,
+//	        "max_prompt_tokens": 168000,
+//	        "max_output_tokens": 32000
+//	    }
+//	}
+func (e *CopilotModelEntry) Limits() *CopilotModelLimits {
+	if e.Capabilities == nil {
+		return nil
+	}
+	limitsRaw, ok := e.Capabilities["limits"]
+	if !ok {
+		return nil
+	}
+	limitsMap, ok := limitsRaw.(map[string]any)
+	if !ok {
+		return nil
+	}
+
+	result := &CopilotModelLimits{
+		MaxContextWindowTokens: anyToInt(limitsMap["max_context_window_tokens"]),
+		MaxPromptTokens:        anyToInt(limitsMap["max_prompt_tokens"]),
+		MaxOutputTokens:        anyToInt(limitsMap["max_output_tokens"]),
+	}
+
+	// Only return if at least one field is populated.
+	if result.MaxContextWindowTokens == 0 && result.MaxPromptTokens == 0 && result.MaxOutputTokens == 0 {
+		return nil
+	}
+	return result
+}
+
+// anyToInt converts a JSON-decoded numeric value to int.
+// Go's encoding/json decodes numbers into float64 when the target is any/interface{}.
+func anyToInt(v any) int {
+	switch n := v.(type) {
+	case float64:
+		return int(n)
+	case float32:
+		return int(n)
+	case int:
+		return n
+	case int64:
+		return int(n)
+	default:
+		return 0
+	}
+}
+
 // CopilotModelsResponse represents the response from the Copilot /models endpoint.
 type CopilotModelsResponse struct {
 	Data   []CopilotModelEntry `json:"data"`

internal/config/config.go

@@ -211,6 +211,10 @@ type QuotaExceeded struct {
 
 	// SwitchPreviewModel indicates whether to automatically switch to a preview model when a quota is exceeded.
 	SwitchPreviewModel bool `yaml:"switch-preview-model" json:"switch-preview-model"`
+
+	// AntigravityCredits indicates whether to retry Antigravity quota_exhausted 429s once
+	// on the same credential with enabledCreditTypes=["GOOGLE_ONE_AI"].
+	AntigravityCredits bool `yaml:"antigravity-credits" json:"antigravity-credits"`
 }
 
 // RoutingConfig configures how credentials are selected for requests.
@@ -257,8 +261,8 @@ type AmpCode struct {
 	UpstreamAPIKey string `yaml:"upstream-api-key" json:"upstream-api-key"`
 
 	// UpstreamAPIKeys maps client API keys (from top-level api-keys) to upstream API keys.
-	// When a client authenticates with a key that matches an entry, that upstream key is used.
-	// If no match is found, falls back to UpstreamAPIKey (default behavior).
+	// When a request is authenticated with one of the APIKeys, the corresponding UpstreamAPIKey
+	// is used for the upstream Amp request.
 	UpstreamAPIKeys []AmpUpstreamAPIKeyEntry `yaml:"upstream-api-keys,omitempty" json:"upstream-api-keys,omitempty"`
 
 	// RestrictManagementToLocalhost restricts Amp management routes (/api/user, /api/threads, etc.)
@@ -380,6 +384,11 @@ type ClaudeKey struct {
 
 	// Cloak configures request cloaking for non-Claude-Code clients.
 	Cloak *CloakConfig `yaml:"cloak,omitempty" json:"cloak,omitempty"`
+
+	// ExperimentalCCHSigning enables opt-in final-body cch signing for cloaked
+	// Claude /v1/messages requests. It is disabled by default so upstream seed
+	// changes do not alter the proxy's legacy behavior.
+	ExperimentalCCHSigning bool `yaml:"experimental-cch-signing,omitempty" json:"experimental-cch-signing,omitempty"`
 }
 
 func (k ClaudeKey) GetAPIKey() string  { return k.APIKey }

internal/config/sdk_config.go

@@ -9,6 +9,10 @@ type SDKConfig struct {
 	// ProxyURL is the URL of an optional proxy server to use for outbound requests.
 	ProxyURL string `yaml:"proxy-url" json:"proxy-url"`
 
+	// EnableGeminiCLIEndpoint controls whether Gemini CLI internal endpoints (/v1internal:*) are enabled.
+	// Default is false for safety; when false, /v1internal:* requests are rejected.
+	EnableGeminiCLIEndpoint bool `yaml:"enable-gemini-cli-endpoint" json:"enable-gemini-cli-endpoint"`
+
 	// ForceModelPrefix requires explicit model prefixes (e.g., "teamA/gemini-3-pro-preview")
 	// to target prefixed credentials. When false, unprefixed model requests may use prefixed
 	// credentials as well.

internal/logging/request_logger.go

@@ -4,6 +4,7 @@
 package logging
 
 import (
+	"bufio"
 	"bytes"
 	"compress/flate"
 	"compress/gzip"
@@ -41,15 +42,17 @@ type RequestLogger interface {
 	//   - statusCode: The response status code
 	//   - responseHeaders: The response headers
 	//   - response: The raw response data
+	//   - websocketTimeline: Optional downstream websocket event timeline
 	//   - apiRequest: The API request data
 	//   - apiResponse: The API response data
+	//   - apiWebsocketTimeline: Optional upstream websocket event timeline
 	//   - requestID: Optional request ID for log file naming
 	//   - requestTimestamp: When the request was received
 	//   - apiResponseTimestamp: When the API response was received
 	//
 	// Returns:
 	//   - error: An error if logging fails, nil otherwise
-	LogRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, apiRequest, apiResponse []byte, apiResponseErrors []*interfaces.ErrorMessage, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error
+	LogRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline []byte, apiResponseErrors []*interfaces.ErrorMessage, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error
 
 	// LogStreamingRequest initiates logging for a streaming request and returns a writer for chunks.
 	//
@@ -111,6 +114,16 @@ type StreamingLogWriter interface {
 	//   - error: An error if writing fails, nil otherwise
 	WriteAPIResponse(apiResponse []byte) error
 
+	// WriteAPIWebsocketTimeline writes the upstream websocket timeline to the log.
+	// This should be called when upstream communication happened over websocket.
+	//
+	// Parameters:
+	//   - apiWebsocketTimeline: The upstream websocket event timeline
+	//
+	// Returns:
+	//   - error: An error if writing fails, nil otherwise
+	WriteAPIWebsocketTimeline(apiWebsocketTimeline []byte) error
+
 	// SetFirstChunkTimestamp sets the TTFB timestamp captured when first chunk was received.
 	//
 	// Parameters:
@@ -203,17 +216,17 @@ func (l *FileRequestLogger) SetErrorLogsMaxFiles(maxFiles int) {
 //
 // Returns:
 //   - error: An error if logging fails, nil otherwise
-func (l *FileRequestLogger) LogRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, apiRequest, apiResponse []byte, apiResponseErrors []*interfaces.ErrorMessage, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
-	return l.logRequest(url, method, requestHeaders, body, statusCode, responseHeaders, response, apiRequest, apiResponse, apiResponseErrors, false, requestID, requestTimestamp, apiResponseTimestamp)
+func (l *FileRequestLogger) LogRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline []byte, apiResponseErrors []*interfaces.ErrorMessage, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
+	return l.logRequest(url, method, requestHeaders, body, statusCode, responseHeaders, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline, apiResponseErrors, false, requestID, requestTimestamp, apiResponseTimestamp)
 }
 
 // LogRequestWithOptions logs a request with optional forced logging behavior.
 // The force flag allows writing error logs even when regular request logging is disabled.
-func (l *FileRequestLogger) LogRequestWithOptions(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, apiRequest, apiResponse []byte, apiResponseErrors []*interfaces.ErrorMessage, force bool, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
-	return l.logRequest(url, method, requestHeaders, body, statusCode, responseHeaders, response, apiRequest, apiResponse, apiResponseErrors, force, requestID, requestTimestamp, apiResponseTimestamp)
+func (l *FileRequestLogger) LogRequestWithOptions(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline []byte, apiResponseErrors []*interfaces.ErrorMessage, force bool, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
+	return l.logRequest(url, method, requestHeaders, body, statusCode, responseHeaders, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline, apiResponseErrors, force, requestID, requestTimestamp, apiResponseTimestamp)
 }
 
-func (l *FileRequestLogger) logRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, apiRequest, apiResponse []byte, apiResponseErrors []*interfaces.ErrorMessage, force bool, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
+func (l *FileRequestLogger) logRequest(url, method string, requestHeaders map[string][]string, body []byte, statusCode int, responseHeaders map[string][]string, response, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline []byte, apiResponseErrors []*interfaces.ErrorMessage, force bool, requestID string, requestTimestamp, apiResponseTimestamp time.Time) error {
 	if !l.enabled && !force {
 		return nil
 	}
@@ -260,8 +273,10 @@ func (l *FileRequestLogger) logRequest(url, method string, requestHeaders map[st
 		requestHeaders,
 		body,
 		requestBodyPath,
+		websocketTimeline,
 		apiRequest,
 		apiResponse,
+		apiWebsocketTimeline,
 		apiResponseErrors,
 		statusCode,
 		responseHeaders,
@@ -518,8 +533,10 @@ func (l *FileRequestLogger) writeNonStreamingLog(
 	requestHeaders map[string][]string,
 	requestBody []byte,
 	requestBodyPath string,
+	websocketTimeline []byte,
 	apiRequest []byte,
 	apiResponse []byte,
+	apiWebsocketTimeline []byte,
 	apiResponseErrors []*interfaces.ErrorMessage,
 	statusCode int,
 	responseHeaders map[string][]string,
@@ -531,7 +548,16 @@ func (l *FileRequestLogger) writeNonStreamingLog(
 	if requestTimestamp.IsZero() {
 		requestTimestamp = time.Now()
 	}
-	if errWrite := writeRequestInfoWithBody(w, url, method, requestHeaders, requestBody, requestBodyPath, requestTimestamp); errWrite != nil {
+	isWebsocketTranscript := hasSectionPayload(websocketTimeline)
+	downstreamTransport := inferDownstreamTransport(requestHeaders, websocketTimeline)
+	upstreamTransport := inferUpstreamTransport(apiRequest, apiResponse, apiWebsocketTimeline, apiResponseErrors)
+	if errWrite := writeRequestInfoWithBody(w, url, method, requestHeaders, requestBody, requestBodyPath, requestTimestamp, downstreamTransport, upstreamTransport, !isWebsocketTranscript); errWrite != nil {
+		return errWrite
+	}
+	if errWrite := writeAPISection(w, "=== WEBSOCKET TIMELINE ===\n", "=== WEBSOCKET TIMELINE", websocketTimeline, time.Time{}); errWrite != nil {
+		return errWrite
+	}
+	if errWrite := writeAPISection(w, "=== API WEBSOCKET TIMELINE ===\n", "=== API WEBSOCKET TIMELINE", apiWebsocketTimeline, time.Time{}); errWrite != nil {
 		return errWrite
 	}
 	if errWrite := writeAPISection(w, "=== API REQUEST ===\n", "=== API REQUEST", apiRequest, time.Time{}); errWrite != nil {
@@ -543,6 +569,12 @@ func (l *FileRequestLogger) writeNonStreamingLog(
 	if errWrite := writeAPISection(w, "=== API RESPONSE ===\n", "=== API RESPONSE", apiResponse, apiResponseTimestamp); errWrite != nil {
 		return errWrite
 	}
+	if isWebsocketTranscript {
+		// Intentionally omit the generic downstream HTTP response section for websocket
+		// transcripts. The durable session exchange is captured in WEBSOCKET TIMELINE,
+		// and appending a one-off upgrade response snapshot would dilute that transcript.
+		return nil
+	}
 	return writeResponseSection(w, statusCode, true, responseHeaders, bytes.NewReader(response), decompressErr, true)
 }
 
@@ -553,6 +585,9 @@ func writeRequestInfoWithBody(
 	body []byte,
 	bodyPath string,
 	timestamp time.Time,
+	downstreamTransport string,
+	upstreamTransport string,
+	includeBody bool,
 ) error {
 	if _, errWrite := io.WriteString(w, "=== REQUEST INFO ===\n"); errWrite != nil {
 		return errWrite
@@ -566,10 +601,20 @@ func writeRequestInfoWithBody(
 	if _, errWrite := io.WriteString(w, fmt.Sprintf("Method: %s\n", method)); errWrite != nil {
 		return errWrite
 	}
+	if strings.TrimSpace(downstreamTransport) != "" {
+		if _, errWrite := io.WriteString(w, fmt.Sprintf("Downstream Transport: %s\n", downstreamTransport)); errWrite != nil {
+			return errWrite
+		}
+	}
+	if strings.TrimSpace(upstreamTransport) != "" {
+		if _, errWrite := io.WriteString(w, fmt.Sprintf("Upstream Transport: %s\n", upstreamTransport)); errWrite != nil {
+			return errWrite
+		}
+	}
 	if _, errWrite := io.WriteString(w, fmt.Sprintf("Timestamp: %s\n", timestamp.Format(time.RFC3339Nano))); errWrite != nil {
 		return errWrite
 	}
-	if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
+	if errWrite := writeSectionSpacing(w, 1); errWrite != nil {
 		return errWrite
 	}
 
@@ -584,36 +629,121 @@ func writeRequestInfoWithBody(
 			}
 		}
 	}
-	if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
+	if errWrite := writeSectionSpacing(w, 1); errWrite != nil {
 		return errWrite
 	}
 
+	if !includeBody {
+		return nil
+	}
+
 	if _, errWrite := io.WriteString(w, "=== REQUEST BODY ===\n"); errWrite != nil {
 		return errWrite
 	}
 
+	bodyTrailingNewlines := 1
 	if bodyPath != "" {
 		bodyFile, errOpen := os.Open(bodyPath)
 		if errOpen != nil {
 			return errOpen
 		}
-		if _, errCopy := io.Copy(w, bodyFile); errCopy != nil {
+		tracker := &trailingNewlineTrackingWriter{writer: w}
+		written, errCopy := io.Copy(tracker, bodyFile)
+		if errCopy != nil {
 			_ = bodyFile.Close()
 			return errCopy
 		}
+		if written > 0 {
+			bodyTrailingNewlines = tracker.trailingNewlines
+		}
 		if errClose := bodyFile.Close(); errClose != nil {
 			log.WithError(errClose).Warn("failed to close request body temp file")
 		}
 	} else if _, errWrite := w.Write(body); errWrite != nil {
 		return errWrite
+	} else if len(body) > 0 {
+		bodyTrailingNewlines = countTrailingNewlinesBytes(body)
 	}
-
-	if _, errWrite := io.WriteString(w, "\n\n"); errWrite != nil {
+	if errWrite := writeSectionSpacing(w, bodyTrailingNewlines); errWrite != nil {
 		return errWrite
 	}
 	return nil
 }
 
+func countTrailingNewlinesBytes(payload []byte) int {
+	count := 0
+	for i := len(payload) - 1; i >= 0; i-- {
+		if payload[i] != '\n' {
+			break
+		}
+		count++
+	}
+	return count
+}
+
+func writeSectionSpacing(w io.Writer, trailingNewlines int) error {
+	missingNewlines := 3 - trailingNewlines
+	if missingNewlines <= 0 {
+		return nil
+	}
+	_, errWrite := io.WriteString(w, strings.Repeat("\n", missingNewlines))
+	return errWrite
+}
+
+type trailingNewlineTrackingWriter struct {
+	writer           io.Writer
+	trailingNewlines int
+}
+
+func (t *trailingNewlineTrackingWriter) Write(payload []byte) (int, error) {
+	written, errWrite := t.writer.Write(payload)
+	if written > 0 {
+		writtenPayload := payload[:written]
+		trailingNewlines := countTrailingNewlinesBytes(writtenPayload)
+		if trailingNewlines == len(writtenPayload) {
+			t.trailingNewlines += trailingNewlines
+		} else {
+			t.trailingNewlines = trailingNewlines
+		}
+	}
+	return written, errWrite
+}
+
+func hasSectionPayload(payload []byte) bool {
+	return len(bytes.TrimSpace(payload)) > 0
+}
+
+func inferDownstreamTransport(headers map[string][]string, websocketTimeline []byte) string {
+	if hasSectionPayload(websocketTimeline) {
+		return "websocket"
+	}
+	for key, values := range headers {
+		if strings.EqualFold(strings.TrimSpace(key), "Upgrade") {
+			for _, value := range values {
+				if strings.EqualFold(strings.TrimSpace(value), "websocket") {
+					return "websocket"
+				}
+			}
+		}
+	}
+	return "http"
+}
+
+func inferUpstreamTransport(apiRequest, apiResponse, apiWebsocketTimeline []byte, _ []*interfaces.ErrorMessage) string {
+	hasHTTP := hasSectionPayload(apiRequest) || hasSectionPayload(apiResponse)
+	hasWS := hasSectionPayload(apiWebsocketTimeline)
+	switch {
+	case hasHTTP && hasWS:
+		return "websocket+http"
+	case hasWS:
+		return "websocket"
+	case hasHTTP:
+		return "http"
+	default:
+		return ""
+	}
+}
+
 func writeAPISection(w io.Writer, sectionHeader string, sectionPrefix string, payload []byte, timestamp time.Time) error {
 	if len(payload) == 0 {
 		return nil
@@ -623,11 +753,6 @@ func writeAPISection(w io.Writer, sectionHeader string, sectionPrefix string, pa
 		if _, errWrite := w.Write(payload); errWrite != nil {
 			return errWrite
 		}
-		if !bytes.HasSuffix(payload, []byte("\n")) {
-			if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
-				return errWrite
-			}
-		}
 	} else {
 		if _, errWrite := io.WriteString(w, sectionHeader); errWrite != nil {
 			return errWrite
@@ -640,12 +765,9 @@ func writeAPISection(w io.Writer, sectionHeader string, sectionPrefix string, pa
 		if _, errWrite := w.Write(payload); errWrite != nil {
 			return errWrite
 		}
-		if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
-			return errWrite
-		}
 	}
 
-	if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
+	if errWrite := writeSectionSpacing(w, countTrailingNewlinesBytes(payload)); errWrite != nil {
 		return errWrite
 	}
 	return nil
@@ -662,12 +784,17 @@ func writeAPIErrorResponses(w io.Writer, apiResponseErrors []*interfaces.ErrorMe
 		if _, errWrite := io.WriteString(w, fmt.Sprintf("HTTP Status: %d\n", apiResponseErrors[i].StatusCode)); errWrite != nil {
 			return errWrite
 		}
+		trailingNewlines := 1
 		if apiResponseErrors[i].Error != nil {
-			if _, errWrite := io.WriteString(w, apiResponseErrors[i].Error.Error()); errWrite != nil {
+			errText := apiResponseErrors[i].Error.Error()
+			if _, errWrite := io.WriteString(w, errText); errWrite != nil {
 				return errWrite
 			}
+			if errText != "" {
+				trailingNewlines = countTrailingNewlinesBytes([]byte(errText))
+			}
 		}
-		if _, errWrite := io.WriteString(w, "\n\n"); errWrite != nil {
+		if errWrite := writeSectionSpacing(w, trailingNewlines); errWrite != nil {
 			return errWrite
 		}
 	}
@@ -694,12 +821,18 @@ func writeResponseSection(w io.Writer, statusCode int, statusWritten bool, respo
 		}
 	}
 
-	if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
-		return errWrite
+	var bufferedReader *bufio.Reader
+	if responseReader != nil {
+		bufferedReader = bufio.NewReader(responseReader)
+	}
+	if !responseBodyStartsWithLeadingNewline(bufferedReader) {
+		if _, errWrite := io.WriteString(w, "\n"); errWrite != nil {
+			return errWrite
+		}
 	}
 
-	if responseReader != nil {
-		if _, errCopy := io.Copy(w, responseReader); errCopy != nil {
+	if bufferedReader != nil {
+		if _, errCopy := io.Copy(w, bufferedReader); errCopy != nil {
 			return errCopy
 		}
 	}
@@ -717,6 +850,19 @@ func writeResponseSection(w io.Writer, statusCode int, statusWritten bool, respo
 	return nil
 }
 
+func responseBodyStartsWithLeadingNewline(reader *bufio.Reader) bool {
+	if reader == nil {
+		return false
+	}
+	if peeked, _ := reader.Peek(2); len(peeked) >= 2 && peeked[0] == '\r' && peeked[1] == '\n' {
+		return true
+	}
+	if peeked, _ := reader.Peek(1); len(peeked) >= 1 && peeked[0] == '\n' {
+		return true
+	}
+	return false
+}
+
 // formatLogContent creates the complete log content for non-streaming requests.
 //
 // Parameters:
@@ -724,6 +870,7 @@ func writeResponseSection(w io.Writer, statusCode int, statusWritten bool, respo
 //   - method: The HTTP method
 //   - headers: The request headers
 //   - body: The request body
+//   - websocketTimeline: The downstream websocket event timeline
 //   - apiRequest: The API request data
 //   - apiResponse: The API response data
 //   - response: The raw response data
@@ -732,11 +879,42 @@ func writeResponseSection(w io.Writer, statusCode int, statusWritten bool, respo
 //
 // Returns:
 //   - string: The formatted log content
-func (l *FileRequestLogger) formatLogContent(url, method string, headers map[string][]string, body, apiRequest, apiResponse, response []byte, status int, responseHeaders map[string][]string, apiResponseErrors []*interfaces.ErrorMessage) string {
+func (l *FileRequestLogger) formatLogContent(url, method string, headers map[string][]string, body, websocketTimeline, apiRequest, apiResponse, apiWebsocketTimeline, response []byte, status int, responseHeaders map[string][]string, apiResponseErrors []*interfaces.ErrorMessage) string {
 	var content strings.Builder
+	isWebsocketTranscript := hasSectionPayload(websocketTimeline)
+	downstreamTransport := inferDownstreamTransport(headers, websocketTimeline)
+	upstreamTransport := inferUpstreamTransport(apiRequest, apiResponse, apiWebsocketTimeline, apiResponseErrors)
 
 	// Request info
-	content.WriteString(l.formatRequestInfo(url, method, headers, body))
+	content.WriteString(l.formatRequestInfo(url, method, headers, body, downstreamTransport, upstreamTransport, !isWebsocketTranscript))
+
+	if len(websocketTimeline) > 0 {
+		if bytes.HasPrefix(websocketTimeline, []byte("=== WEBSOCKET TIMELINE")) {
+			content.Write(websocketTimeline)
+			if !bytes.HasSuffix(websocketTimeline, []byte("\n")) {
+				content.WriteString("\n")
+			}
+		} else {
+			content.WriteString("=== WEBSOCKET TIMELINE ===\n")
+			content.Write(websocketTimeline)
+			content.WriteString("\n")
+		}
+		content.WriteString("\n")
+	}
+
+	if len(apiWebsocketTimeline) > 0 {
+		if bytes.HasPrefix(apiWebsocketTimeline, []byte("=== API WEBSOCKET TIMELINE")) {
+			content.Write(apiWebsocketTimeline)
+			if !bytes.HasSuffix(apiWebsocketTimeline, []byte("\n")) {
+				content.WriteString("\n")
+			}
+		} else {
+			content.WriteString("=== API WEBSOCKET TIMELINE ===\n")
+			content.Write(apiWebsocketTimeline)
+			content.WriteString("\n")
+		}
+		content.WriteString("\n")
+	}
 
 	if len(apiRequest) > 0 {
 		if bytes.HasPrefix(apiRequest, []byte("=== API REQUEST")) {
@@ -773,6 +951,12 @@ func (l *FileRequestLogger) formatLogContent(url, method string, headers map[str
 		content.WriteString("\n")
 	}
 
+	if isWebsocketTranscript {
+		// Mirror writeNonStreamingLog: websocket transcripts end with the dedicated
+		// timeline sections instead of a generic downstream HTTP response block.
+		return content.String()
+	}
+
 	// Response section
 	content.WriteString("=== RESPONSE ===\n")
 	content.WriteString(fmt.Sprintf("Status: %d\n", status))
@@ -933,13 +1117,19 @@ func (l *FileRequestLogger) decompressZstd(data []byte) ([]byte, error) {
 //
 // Returns:
 //   - string: The formatted request information
-func (l *FileRequestLogger) formatRequestInfo(url, method string, headers map[string][]string, body []byte) string {
+func (l *FileRequestLogger) formatRequestInfo(url, method string, headers map[string][]string, body []byte, downstreamTransport string, upstreamTransport string, includeBody bool) string {
 	var content strings.Builder
 
 	content.WriteString("=== REQUEST INFO ===\n")
 	content.WriteString(fmt.Sprintf("Version: %s\n", buildinfo.Version))
 	content.WriteString(fmt.Sprintf("URL: %s\n", url))
 	content.WriteString(fmt.Sprintf("Method: %s\n", method))
+	if strings.TrimSpace(downstreamTransport) != "" {
+		content.WriteString(fmt.Sprintf("Downstream Transport: %s\n", downstreamTransport))
+	}
+	if strings.TrimSpace(upstreamTransport) != "" {
+		content.WriteString(fmt.Sprintf("Upstream Transport: %s\n", upstreamTransport))
+	}
 	content.WriteString(fmt.Sprintf("Timestamp: %s\n", time.Now().Format(time.RFC3339Nano)))
 	content.WriteString("\n")
 
@@ -952,6 +1142,10 @@ func (l *FileRequestLogger) formatRequestInfo(url, method string, headers map[st
 	}
 	content.WriteString("\n")
 
+	if !includeBody {
+		return content.String()
+	}
+
 	content.WriteString("=== REQUEST BODY ===\n")
 	content.Write(body)
 	content.WriteString("\n\n")
@@ -1011,6 +1205,9 @@ type FileStreamingLogWriter struct {
 	// apiResponse stores the upstream API response data.
 	apiResponse []byte
 
+	// apiWebsocketTimeline stores the upstream websocket event timeline.
+	apiWebsocketTimeline []byte
+
 	// apiResponseTimestamp captures when the API response was received.
 	apiResponseTimestamp time.Time
 }
@@ -1092,6 +1289,21 @@ func (w *FileStreamingLogWriter) WriteAPIResponse(apiResponse []byte) error {
 	return nil
 }
 
+// WriteAPIWebsocketTimeline buffers the upstream websocket timeline for later writing.
+//
+// Parameters:
+//   - apiWebsocketTimeline: The upstream websocket event timeline
+//
+// Returns:
+//   - error: Always returns nil (buffering cannot fail)
+func (w *FileStreamingLogWriter) WriteAPIWebsocketTimeline(apiWebsocketTimeline []byte) error {
+	if len(apiWebsocketTimeline) == 0 {
+		return nil
+	}
+	w.apiWebsocketTimeline = bytes.Clone(apiWebsocketTimeline)
+	return nil
+}
+
 func (w *FileStreamingLogWriter) SetFirstChunkTimestamp(timestamp time.Time) {
 	if !timestamp.IsZero() {
 		w.apiResponseTimestamp = timestamp
@@ -1100,7 +1312,7 @@ func (w *FileStreamingLogWriter) SetFirstChunkTimestamp(timestamp time.Time) {
 
 // Close finalizes the log file and cleans up resources.
 // It writes all buffered data to the file in the correct order:
-// API REQUEST -> API RESPONSE -> RESPONSE (status, headers, body chunks)
+// API WEBSOCKET TIMELINE -> API REQUEST -> API RESPONSE -> RESPONSE (status, headers, body chunks)
 //
 // Returns:
 //   - error: An error if closing fails, nil otherwise
@@ -1182,7 +1394,10 @@ func (w *FileStreamingLogWriter) asyncWriter() {
 }
 
 func (w *FileStreamingLogWriter) writeFinalLog(logFile *os.File) error {
-	if errWrite := writeRequestInfoWithBody(logFile, w.url, w.method, w.requestHeaders, nil, w.requestBodyPath, w.timestamp); errWrite != nil {
+	if errWrite := writeRequestInfoWithBody(logFile, w.url, w.method, w.requestHeaders, nil, w.requestBodyPath, w.timestamp, "http", inferUpstreamTransport(w.apiRequest, w.apiResponse, w.apiWebsocketTimeline, nil), true); errWrite != nil {
+		return errWrite
+	}
+	if errWrite := writeAPISection(logFile, "=== API WEBSOCKET TIMELINE ===\n", "=== API WEBSOCKET TIMELINE", w.apiWebsocketTimeline, time.Time{}); errWrite != nil {
 		return errWrite
 	}
 	if errWrite := writeAPISection(logFile, "=== API REQUEST ===\n", "=== API REQUEST", w.apiRequest, time.Time{}); errWrite != nil {
@@ -1265,6 +1480,17 @@ func (w *NoOpStreamingLogWriter) WriteAPIResponse(_ []byte) error {
 	return nil
 }
 
+// WriteAPIWebsocketTimeline is a no-op implementation that does nothing and always returns nil.
+//
+// Parameters:
+//   - apiWebsocketTimeline: The upstream websocket event timeline (ignored)
+//
+// Returns:
+//   - error: Always returns nil
+func (w *NoOpStreamingLogWriter) WriteAPIWebsocketTimeline(_ []byte) error {
+	return nil
+}
+
 func (w *NoOpStreamingLogWriter) SetFirstChunkTimestamp(_ time.Time) {}
 
 // Close is a no-op implementation that does nothing and always returns nil.

internal/registry/model_definitions.go

@@ -93,6 +93,30 @@ func GetAntigravityModels() []*ModelInfo {
 func GetCodeBuddyModels() []*ModelInfo {
 	now := int64(1748044800) // 2025-05-24
 	return []*ModelInfo{
+		{
+			ID:                  "auto",
+			Object:              "model",
+			Created:             now,
+			OwnedBy:             "tencent",
+			Type:                "codebuddy",
+			DisplayName:         "Auto",
+			Description:         "Automatic model selection via CodeBuddy",
+			ContextLength:       128000,
+			MaxCompletionTokens: 32768,
+			SupportedEndpoints:  []string{"/chat/completions"},
+		},
+		{
+			ID:                  "glm-5.0-turbo",
+			Object:              "model",
+			Created:             now,
+			OwnedBy:             "tencent",
+			Type:                "codebuddy",
+			DisplayName:         "GLM-5.0 Turbo",
+			Description:         "GLM-5.0 Turbo via CodeBuddy",
+			ContextLength:       128000,
+			MaxCompletionTokens: 32768,
+			SupportedEndpoints:  []string{"/chat/completions"},
+		},
 		{
 			ID:                  "glm-5.0",
 			Object:              "model",
@@ -118,13 +142,13 @@ func GetCodeBuddyModels() []*ModelInfo {
 			SupportedEndpoints:  []string{"/chat/completions"},
 		},
 		{
-			ID:                  "minimax-m2.5",
+			ID:                  "minimax-m2.7",
 			Object:              "model",
 			Created:             now,
 			OwnedBy:             "tencent",
 			Type:                "codebuddy",
-			DisplayName:         "MiniMax M2.5",
-			Description:         "MiniMax M2.5 via CodeBuddy",
+			DisplayName:         "MiniMax M2.7",
+			Description:         "MiniMax M2.7 via CodeBuddy",
 			ContextLength:       200000,
 			MaxCompletionTokens: 32768,
 			SupportedEndpoints:  []string{"/chat/completions"},
@@ -141,6 +165,19 @@ func GetCodeBuddyModels() []*ModelInfo {
 			MaxCompletionTokens: 32768,
 			SupportedEndpoints:  []string{"/chat/completions"},
 		},
+		{
+			ID:                  "kimi-k2-thinking",
+			Object:              "model",
+			Created:             now,
+			OwnedBy:             "tencent",
+			Type:                "codebuddy",
+			DisplayName:         "Kimi K2 Thinking",
+			Description:         "Kimi K2 Thinking via CodeBuddy",
+			ContextLength:       128000,
+			MaxCompletionTokens: 32768,
+			Thinking:            &ThinkingSupport{ZeroAllowed: true},
+			SupportedEndpoints:  []string{"/chat/completions"},
+		},
 		{
 			ID:                  "deepseek-v3-2-volc",
 			Object:              "model",
@@ -148,24 +185,11 @@ func GetCodeBuddyModels() []*ModelInfo {
 			OwnedBy:             "tencent",
 			Type:                "codebuddy",
 			DisplayName:         "DeepSeek V3.2 (Volc)",
-			Description:         "DeepSeek V3.2 via CodeBuddy (Volcano Engine)",
+			Description:         "DeepSeek V3.2 via CodeBuddy",
 			ContextLength:       128000,
 			MaxCompletionTokens: 32768,
 			SupportedEndpoints:  []string{"/chat/completions"},
 		},
-		{
-			ID:                  "hunyuan-2.0-thinking",
-			Object:              "model",
-			Created:             now,
-			OwnedBy:             "tencent",
-			Type:                "codebuddy",
-			DisplayName:         "Hunyuan 2.0 Thinking",
-			Description:         "Tencent Hunyuan 2.0 Thinking via CodeBuddy",
-			ContextLength:       128000,
-			MaxCompletionTokens: 32768,
-			Thinking:            &ThinkingSupport{ZeroAllowed: true},
-			SupportedEndpoints:  []string{"/chat/completions"},
-		},
 	}
 }
 
@@ -477,6 +501,19 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			SupportedEndpoints:  []string{"/responses"},
 			Thinking:            &ThinkingSupport{Levels: []string{"none", "low", "medium", "high", "xhigh"}},
 		},
+		{
+			ID:                  "gpt-5.4-mini",
+			Object:              "model",
+			Created:             now,
+			OwnedBy:             "github-copilot",
+			Type:                "github-copilot",
+			DisplayName:         "GPT-5.4 mini",
+			Description:         "OpenAI GPT-5.4 mini via GitHub Copilot",
+			ContextLength:       200000,
+			MaxCompletionTokens: 32768,
+			SupportedEndpoints:  []string{"/responses"},
+			Thinking:            &ThinkingSupport{Levels: []string{"none", "low", "medium", "high", "xhigh"}},
+		},
 		{
 			ID:                  "claude-haiku-4.5",
 			Object:              "model",
@@ -512,6 +549,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			ContextLength:       200000,
 			MaxCompletionTokens: 64000,
 			SupportedEndpoints:  []string{"/chat/completions"},
+			Thinking:            &ThinkingSupport{Levels: []string{"low", "medium", "high"}},
 		},
 		{
 			ID:                  "claude-opus-4.6",
@@ -524,6 +562,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			ContextLength:       200000,
 			MaxCompletionTokens: 64000,
 			SupportedEndpoints:  []string{"/chat/completions"},
+			Thinking:            &ThinkingSupport{Levels: []string{"low", "medium", "high"}},
 		},
 		{
 			ID:                  "claude-sonnet-4",
@@ -536,6 +575,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			ContextLength:       200000,
 			MaxCompletionTokens: 64000,
 			SupportedEndpoints:  []string{"/chat/completions"},
+			Thinking:            &ThinkingSupport{Levels: []string{"low", "medium", "high"}},
 		},
 		{
 			ID:                  "claude-sonnet-4.5",
@@ -548,6 +588,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			ContextLength:       200000,
 			MaxCompletionTokens: 64000,
 			SupportedEndpoints:  []string{"/chat/completions"},
+			Thinking:            &ThinkingSupport{Levels: []string{"low", "medium", "high"}},
 		},
 		{
 			ID:                  "claude-sonnet-4.6",
@@ -560,6 +601,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			ContextLength:       200000,
 			MaxCompletionTokens: 64000,
 			SupportedEndpoints:  []string{"/chat/completions"},
+			Thinking:            &ThinkingSupport{Levels: []string{"low", "medium", "high"}},
 		},
 		{
 			ID:                  "gemini-2.5-pro",
@@ -571,6 +613,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			Description:         "Google Gemini 2.5 Pro via GitHub Copilot",
 			ContextLength:       1048576,
 			MaxCompletionTokens: 65536,
+			SupportedEndpoints:  []string{"/chat/completions"},
 		},
 		{
 			ID:                  "gemini-3-pro-preview",
@@ -582,6 +625,7 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			Description:         "Google Gemini 3 Pro Preview via GitHub Copilot",
 			ContextLength:       1048576,
 			MaxCompletionTokens: 65536,
+			SupportedEndpoints:  []string{"/chat/completions"},
 		},
 		{
 			ID:                  "gemini-3.1-pro-preview",
@@ -591,8 +635,9 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			Type:                "github-copilot",
 			DisplayName:         "Gemini 3.1 Pro (Preview)",
 			Description:         "Google Gemini 3.1 Pro Preview via GitHub Copilot",
-			ContextLength:       1048576,
+			ContextLength:       173000,
 			MaxCompletionTokens: 65536,
+			SupportedEndpoints:  []string{"/chat/completions"},
 		},
 		{
 			ID:                  "gemini-3-flash-preview",
@@ -602,8 +647,9 @@ func GetGitHubCopilotModels() []*ModelInfo {
 			Type:                "github-copilot",
 			DisplayName:         "Gemini 3 Flash (Preview)",
 			Description:         "Google Gemini 3 Flash Preview via GitHub Copilot",
-			ContextLength:       1048576,
+			ContextLength:       173000,
 			MaxCompletionTokens: 65536,
+			SupportedEndpoints:  []string{"/chat/completions"},
 		},
 		{
 			ID:                  "grok-code-fast-1",

internal/registry/model_definitions_test.go

@@ -0,0 +1,29 @@
+package registry
+
+import "testing"
+
+func TestGitHubCopilotGeminiModelsAreChatOnly(t *testing.T) {
+	models := GetGitHubCopilotModels()
+	required := map[string]bool{
+		"gemini-2.5-pro":         false,
+		"gemini-3-pro-preview":   false,
+		"gemini-3.1-pro-preview": false,
+		"gemini-3-flash-preview": false,
+	}
+
+	for _, model := range models {
+		if _, ok := required[model.ID]; !ok {
+			continue
+		}
+		required[model.ID] = true
+		if len(model.SupportedEndpoints) != 1 || model.SupportedEndpoints[0] != "/chat/completions" {
+			t.Fatalf("model %q supported endpoints = %v, want [/chat/completions]", model.ID, model.SupportedEndpoints)
+		}
+	}
+
+	for modelID, found := range required {
+		if !found {
+			t.Fatalf("expected GitHub Copilot model %q in definitions", modelID)
+		}
+	}
+}

internal/registry/models/models.json

@@ -280,6 +280,7 @@
         "dynamic_allowed": true,
         "levels": [
           "low",
+          "medium",
           "high"
         ]
       }
@@ -554,6 +555,7 @@
         "dynamic_allowed": true,
         "levels": [
           "low",
+          "medium",
           "high"
         ]
       }
@@ -610,6 +612,8 @@
         "dynamic_allowed": true,
         "levels": [
           "minimal",
+          "low",
+          "medium",
           "high"
         ]
       }
@@ -838,6 +842,7 @@
         "dynamic_allowed": true,
         "levels": [
           "low",
+          "medium",
           "high"
         ]
       }
@@ -896,6 +901,8 @@
         "dynamic_allowed": true,
         "levels": [
           "minimal",
+          "low",
+          "medium",
           "high"
         ]
       }
@@ -1070,6 +1077,8 @@
         "dynamic_allowed": true,
         "levels": [
           "minimal",
+          "low",
+          "medium",
           "high"
         ]
       }
@@ -1371,6 +1380,75 @@
           "xhigh"
         ]
       }
+    },
+    {
+      "id": "gpt-5.3-codex",
+      "object": "model",
+      "created": 1770307200,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.3 Codex",
+      "version": "gpt-5.3",
+      "description": "Stable version of GPT 5.3 Codex, The best model for coding and agentic tasks across domains.",
+      "context_length": 400000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
+    },
+    {
+      "id": "gpt-5.4",
+      "object": "model",
+      "created": 1772668800,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.4",
+      "version": "gpt-5.4",
+      "description": "Stable version of GPT 5.4",
+      "context_length": 1050000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
+    },
+    {
+      "id": "gpt-5.4-mini",
+      "object": "model",
+      "created": 1773705600,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.4 Mini",
+      "version": "gpt-5.4-mini",
+      "description": "GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.",
+      "context_length": 400000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
     }
   ],
   "codex-team": [
@@ -1623,6 +1701,29 @@
           "xhigh"
         ]
       }
+    },
+    {
+      "id": "gpt-5.4-mini",
+      "object": "model",
+      "created": 1773705600,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.4 Mini",
+      "version": "gpt-5.4-mini",
+      "description": "GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.",
+      "context_length": 400000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
     }
   ],
   "codex-plus": [
@@ -1898,6 +1999,29 @@
           "xhigh"
         ]
       }
+    },
+    {
+      "id": "gpt-5.4-mini",
+      "object": "model",
+      "created": 1773705600,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.4 Mini",
+      "version": "gpt-5.4-mini",
+      "description": "GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.",
+      "context_length": 400000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
     }
   ],
   "codex-pro": [
@@ -2173,55 +2297,40 @@
           "xhigh"
         ]
       }
+    },
+    {
+      "id": "gpt-5.4-mini",
+      "object": "model",
+      "created": 1773705600,
+      "owned_by": "openai",
+      "type": "openai",
+      "display_name": "GPT 5.4 Mini",
+      "version": "gpt-5.4-mini",
+      "description": "GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.",
+      "context_length": 400000,
+      "max_completion_tokens": 128000,
+      "supported_parameters": [
+        "tools"
+      ],
+      "thinking": {
+        "levels": [
+          "low",
+          "medium",
+          "high",
+          "xhigh"
+        ]
+      }
     }
   ],
   "qwen": [
-    {
-      "id": "qwen3-coder-plus",
-      "object": "model",
-      "created": 1753228800,
-      "owned_by": "qwen",
-      "type": "qwen",
-      "display_name": "Qwen3 Coder Plus",
-      "version": "3.0",
-      "description": "Advanced code generation and understanding model",
-      "context_length": 32768,
-      "max_completion_tokens": 8192,
-      "supported_parameters": [
-        "temperature",
-        "top_p",
-        "max_tokens",
-        "stream",
-        "stop"
-      ]
-    },
-    {
-      "id": "qwen3-coder-flash",
-      "object": "model",
-      "created": 1753228800,
-      "owned_by": "qwen",
-      "type": "qwen",
-      "display_name": "Qwen3 Coder Flash",
-      "version": "3.0",
-      "description": "Fast code generation model",
-      "context_length": 8192,
-      "max_completion_tokens": 2048,
-      "supported_parameters": [
-        "temperature",
-        "top_p",
-        "max_tokens",
-        "stream",
-        "stop"
-      ]
-    },
     {
       "id": "coder-model",
       "object": "model",
       "created": 1771171200,
       "owned_by": "qwen",
       "type": "qwen",
-      "display_name": "Qwen 3.5 Plus",
-      "version": "3.5",
+      "display_name": "Qwen 3.6 Plus",
+      "version": "3.6",
       "description": "efficient hybrid model with leading coding performance",
       "context_length": 1048576,
       "max_completion_tokens": 65536,
@@ -2232,25 +2341,6 @@
         "stream",
         "stop"
       ]
-    },
-    {
-      "id": "vision-model",
-      "object": "model",
-      "created": 1758672000,
-      "owned_by": "qwen",
-      "type": "qwen",
-      "display_name": "Qwen3 Vision Model",
-      "version": "3.0",
-      "description": "Vision model model",
-      "context_length": 32768,
-      "max_completion_tokens": 2048,
-      "supported_parameters": [
-        "temperature",
-        "top_p",
-        "max_tokens",
-        "stream",
-        "stop"
-      ]
     }
   ],
   "iflow": [
@@ -2639,11 +2729,12 @@
       "context_length": 1048576,
       "max_completion_tokens": 65535,
       "thinking": {
-        "min": 128,
-        "max": 32768,
+        "min": 1,
+        "max": 65535,
         "dynamic_allowed": true,
         "levels": [
           "low",
+          "medium",
           "high"
         ]
       }
@@ -2659,11 +2750,12 @@
       "context_length": 1048576,
       "max_completion_tokens": 65535,
       "thinking": {
-        "min": 128,
-        "max": 32768,
+        "min": 1,
+        "max": 65535,
         "dynamic_allowed": true,
         "levels": [
           "low",
+          "medium",
           "high"
         ]
       }

internal/runtime/executor/aistudio_executor.go

@@ -14,7 +14,9 @@ import (
 	"strings"
 
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/wsrelay"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
@@ -46,8 +48,16 @@ func NewAIStudioExecutor(cfg *config.Config, provider string, relay *wsrelay.Man
 // Identifier returns the executor identifier.
 func (e *AIStudioExecutor) Identifier() string { return "aistudio" }
 
-// PrepareRequest prepares the HTTP request for execution (no-op for AI Studio).
-func (e *AIStudioExecutor) PrepareRequest(_ *http.Request, _ *cliproxyauth.Auth) error {
+// PrepareRequest prepares the HTTP request for execution.
+func (e *AIStudioExecutor) PrepareRequest(req *http.Request, auth *cliproxyauth.Auth) error {
+	if req == nil {
+		return nil
+	}
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(req, attrs)
 	return nil
 }
 
@@ -66,6 +76,9 @@ func (e *AIStudioExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.A
 		return nil, fmt.Errorf("aistudio executor: missing auth")
 	}
 	httpReq := req.WithContext(ctx)
+	if err := e.PrepareRequest(httpReq, auth); err != nil {
+		return nil, err
+	}
 	if httpReq.URL == nil || strings.TrimSpace(httpReq.URL.String()) == "" {
 		return nil, fmt.Errorf("aistudio executor: request URL is empty")
 	}
@@ -115,8 +128,8 @@ func (e *AIStudioExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth,
 		return resp, statusErr{code: http.StatusNotImplemented, msg: "/responses/compact not supported"}
 	}
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	translatedReq, body, err := e.translateRequest(req, opts, false)
 	if err != nil {
@@ -130,6 +143,11 @@ func (e *AIStudioExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth,
 		Headers: http.Header{"Content-Type": []string{"application/json"}},
 		Body:    body.payload,
 	}
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(&http.Request{Header: wsReq.Headers}, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -137,7 +155,7 @@ func (e *AIStudioExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth,
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       endpoint,
 		Method:    http.MethodPost,
 		Headers:   wsReq.Headers.Clone(),
@@ -151,17 +169,17 @@ func (e *AIStudioExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth,
 
 	wsResp, err := e.relay.NonStream(ctx, authID, wsReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, wsResp.Status, wsResp.Headers.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, wsResp.Status, wsResp.Headers.Clone())
 	if len(wsResp.Body) > 0 {
-		appendAPIResponseChunk(ctx, e.cfg, wsResp.Body)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, wsResp.Body)
 	}
 	if wsResp.Status < 200 || wsResp.Status >= 300 {
 		return resp, statusErr{code: wsResp.Status, msg: string(wsResp.Body)}
 	}
-	reporter.publish(ctx, parseGeminiUsage(wsResp.Body))
+	reporter.Publish(ctx, helps.ParseGeminiUsage(wsResp.Body))
 	var param any
 	out := sdktranslator.TranslateNonStream(ctx, body.toFormat, opts.SourceFormat, req.Model, opts.OriginalRequest, translatedReq, wsResp.Body, &param)
 	resp = cliproxyexecutor.Response{Payload: ensureColonSpacedJSON(out), Headers: wsResp.Headers.Clone()}
@@ -174,8 +192,8 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 		return nil, statusErr{code: http.StatusNotImplemented, msg: "/responses/compact not supported"}
 	}
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	translatedReq, body, err := e.translateRequest(req, opts, true)
 	if err != nil {
@@ -189,13 +207,18 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 		Headers: http.Header{"Content-Type": []string{"application/json"}},
 		Body:    body.payload,
 	}
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(&http.Request{Header: wsReq.Headers}, attrs)
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
 		authID = auth.ID
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       endpoint,
 		Method:    http.MethodPost,
 		Headers:   wsReq.Headers.Clone(),
@@ -208,24 +231,24 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 	})
 	wsStream, err := e.relay.Stream(ctx, authID, wsReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
 	firstEvent, ok := <-wsStream
 	if !ok {
 		err = fmt.Errorf("wsrelay: stream closed before start")
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
 	if firstEvent.Status > 0 && firstEvent.Status != http.StatusOK {
 		metadataLogged := false
 		if firstEvent.Status > 0 {
-			recordAPIResponseMetadata(ctx, e.cfg, firstEvent.Status, firstEvent.Headers.Clone())
+			helps.RecordAPIResponseMetadata(ctx, e.cfg, firstEvent.Status, firstEvent.Headers.Clone())
 			metadataLogged = true
 		}
 		var body bytes.Buffer
 		if len(firstEvent.Payload) > 0 {
-			appendAPIResponseChunk(ctx, e.cfg, firstEvent.Payload)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, firstEvent.Payload)
 			body.Write(firstEvent.Payload)
 		}
 		if firstEvent.Type == wsrelay.MessageTypeStreamEnd {
@@ -233,18 +256,18 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 		}
 		for event := range wsStream {
 			if event.Err != nil {
-				recordAPIResponseError(ctx, e.cfg, event.Err)
+				helps.RecordAPIResponseError(ctx, e.cfg, event.Err)
 				if body.Len() == 0 {
 					body.WriteString(event.Err.Error())
 				}
 				break
 			}
 			if !metadataLogged && event.Status > 0 {
-				recordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
+				helps.RecordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
 				metadataLogged = true
 			}
 			if len(event.Payload) > 0 {
-				appendAPIResponseChunk(ctx, e.cfg, event.Payload)
+				helps.AppendAPIResponseChunk(ctx, e.cfg, event.Payload)
 				body.Write(event.Payload)
 			}
 			if event.Type == wsrelay.MessageTypeStreamEnd {
@@ -260,23 +283,23 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 		metadataLogged := false
 		processEvent := func(event wsrelay.StreamEvent) bool {
 			if event.Err != nil {
-				recordAPIResponseError(ctx, e.cfg, event.Err)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIResponseError(ctx, e.cfg, event.Err)
+				reporter.PublishFailure(ctx)
 				out <- cliproxyexecutor.StreamChunk{Err: fmt.Errorf("wsrelay: %v", event.Err)}
 				return false
 			}
 			switch event.Type {
 			case wsrelay.MessageTypeStreamStart:
 				if !metadataLogged && event.Status > 0 {
-					recordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
+					helps.RecordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
 					metadataLogged = true
 				}
 			case wsrelay.MessageTypeStreamChunk:
 				if len(event.Payload) > 0 {
-					appendAPIResponseChunk(ctx, e.cfg, event.Payload)
-					filtered := FilterSSEUsageMetadata(event.Payload)
-					if detail, ok := parseGeminiStreamUsage(filtered); ok {
-						reporter.publish(ctx, detail)
+					helps.AppendAPIResponseChunk(ctx, e.cfg, event.Payload)
+					filtered := helps.FilterSSEUsageMetadata(event.Payload)
+					if detail, ok := helps.ParseGeminiStreamUsage(filtered); ok {
+						reporter.Publish(ctx, detail)
 					}
 					lines := sdktranslator.TranslateStream(ctx, body.toFormat, opts.SourceFormat, req.Model, opts.OriginalRequest, translatedReq, filtered, &param)
 					for i := range lines {
@@ -288,21 +311,21 @@ func (e *AIStudioExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth
 				return false
 			case wsrelay.MessageTypeHTTPResp:
 				if !metadataLogged && event.Status > 0 {
-					recordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
+					helps.RecordAPIResponseMetadata(ctx, e.cfg, event.Status, event.Headers.Clone())
 					metadataLogged = true
 				}
 				if len(event.Payload) > 0 {
-					appendAPIResponseChunk(ctx, e.cfg, event.Payload)
+					helps.AppendAPIResponseChunk(ctx, e.cfg, event.Payload)
 				}
 				lines := sdktranslator.TranslateStream(ctx, body.toFormat, opts.SourceFormat, req.Model, opts.OriginalRequest, translatedReq, event.Payload, &param)
 				for i := range lines {
 					out <- cliproxyexecutor.StreamChunk{Payload: ensureColonSpacedJSON(lines[i])}
 				}
-				reporter.publish(ctx, parseGeminiUsage(event.Payload))
+				reporter.Publish(ctx, helps.ParseGeminiUsage(event.Payload))
 				return false
 			case wsrelay.MessageTypeError:
-				recordAPIResponseError(ctx, e.cfg, event.Err)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIResponseError(ctx, e.cfg, event.Err)
+				reporter.PublishFailure(ctx)
 				out <- cliproxyexecutor.StreamChunk{Err: fmt.Errorf("wsrelay: %v", event.Err)}
 				return false
 			}
@@ -345,7 +368,7 @@ func (e *AIStudioExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.A
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       endpoint,
 		Method:    http.MethodPost,
 		Headers:   wsReq.Headers.Clone(),
@@ -358,12 +381,12 @@ func (e *AIStudioExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.A
 	})
 	resp, err := e.relay.NonStream(ctx, authID, wsReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return cliproxyexecutor.Response{}, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, resp.Status, resp.Headers.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, resp.Status, resp.Headers.Clone())
 	if len(resp.Body) > 0 {
-		appendAPIResponseChunk(ctx, e.cfg, resp.Body)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, resp.Body)
 	}
 	if resp.Status < 200 || resp.Status >= 300 {
 		return cliproxyexecutor.Response{}, statusErr{code: resp.Status, msg: string(resp.Body)}
@@ -404,8 +427,8 @@ func (e *AIStudioExecutor) translateRequest(req cliproxyexecutor.Request, opts c
 		return nil, translatedPayload{}, err
 	}
 	payload = fixGeminiImageAspectRatio(baseModel, payload)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	payload = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", payload, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	payload = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", payload, originalTranslated, requestedModel)
 	payload, _ = sjson.DeleteBytes(payload, "generationConfig.maxOutputTokens")
 	payload, _ = sjson.DeleteBytes(payload, "generationConfig.responseMimeType")
 	payload, _ = sjson.DeleteBytes(payload, "generationConfig.responseJsonSchema")

internal/runtime/executor/antigravity_executor.go

@@ -24,6 +24,7 @@ import (
 
 	"github.com/google/uuid"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	sdkAuth "github.com/router-for-me/CLIProxyAPI/v6/sdk/auth"
@@ -47,12 +48,41 @@ const (
 	defaultAntigravityAgent        = "antigravity/1.19.6 darwin/arm64"
 	antigravityAuthType            = "antigravity"
 	refreshSkew                    = 3000 * time.Second
+	antigravityCreditsRetryTTL     = 5 * time.Hour
 	// systemInstruction              = "You are Antigravity, a powerful agentic AI coding assistant designed by the Google Deepmind team working on Advanced Agentic Coding.You are pair programming with a USER to solve their coding task. The task may require creating a new codebase, modifying or debugging an existing codebase, or simply answering a question.**Absolute paths only****Proactiveness**"
 )
 
+type antigravity429Category string
+
+const (
+	antigravity429Unknown        antigravity429Category = "unknown"
+	antigravity429RateLimited    antigravity429Category = "rate_limited"
+	antigravity429QuotaExhausted antigravity429Category = "quota_exhausted"
+)
+
 var (
-	randSource      = rand.New(rand.NewSource(time.Now().UnixNano()))
-	randSourceMutex sync.Mutex
+	randSource                        = rand.New(rand.NewSource(time.Now().UnixNano()))
+	randSourceMutex                   sync.Mutex
+	antigravityCreditsExhaustedByAuth sync.Map
+	antigravityPreferCreditsByModel   sync.Map
+	antigravityQuotaExhaustedKeywords = []string{
+		"quota_exhausted",
+		"quota exhausted",
+	}
+	antigravityCreditsExhaustedKeywords = []string{
+		"google_one_ai",
+		"insufficient credit",
+		"insufficient credits",
+		"not enough credit",
+		"not enough credits",
+		"credit exhausted",
+		"credits exhausted",
+		"credit balance",
+		"minimumcreditamountforusage",
+		"minimum credit amount for usage",
+		"minimum credit",
+		"resource has been exhausted",
+	}
 )
 
 // AntigravityExecutor proxies requests to the antigravity upstream.
@@ -113,7 +143,7 @@ func initAntigravityTransport() {
 func newAntigravityHTTPClient(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
 	antigravityTransportOnce.Do(initAntigravityTransport)
 
-	client := newProxyAwareHTTPClient(ctx, cfg, auth, timeout)
+	client := helps.NewProxyAwareHTTPClient(ctx, cfg, auth, timeout)
 	// If no transport is set, use the shared HTTP/1.1 transport.
 	if client.Transport == nil {
 		client.Transport = antigravityTransport
@@ -183,6 +213,231 @@ func (e *AntigravityExecutor) HttpRequest(ctx context.Context, auth *cliproxyaut
 	return httpClient.Do(httpReq)
 }
 
+func injectEnabledCreditTypes(payload []byte) []byte {
+	if len(payload) == 0 {
+		return nil
+	}
+	if !gjson.ValidBytes(payload) {
+		return nil
+	}
+	updated, err := sjson.SetRawBytes(payload, "enabledCreditTypes", []byte(`["GOOGLE_ONE_AI"]`))
+	if err != nil {
+		return nil
+	}
+	return updated
+}
+
+func classifyAntigravity429(body []byte) antigravity429Category {
+	if len(body) == 0 {
+		return antigravity429Unknown
+	}
+	lowerBody := strings.ToLower(string(body))
+	for _, keyword := range antigravityQuotaExhaustedKeywords {
+		if strings.Contains(lowerBody, keyword) {
+			return antigravity429QuotaExhausted
+		}
+	}
+	status := strings.TrimSpace(gjson.GetBytes(body, "error.status").String())
+	if !strings.EqualFold(status, "RESOURCE_EXHAUSTED") {
+		return antigravity429Unknown
+	}
+	details := gjson.GetBytes(body, "error.details")
+	if !details.Exists() || !details.IsArray() {
+		return antigravity429Unknown
+	}
+	for _, detail := range details.Array() {
+		if detail.Get("@type").String() != "type.googleapis.com/google.rpc.ErrorInfo" {
+			continue
+		}
+		reason := strings.TrimSpace(detail.Get("reason").String())
+		if strings.EqualFold(reason, "QUOTA_EXHAUSTED") {
+			return antigravity429QuotaExhausted
+		}
+		if strings.EqualFold(reason, "RATE_LIMIT_EXCEEDED") {
+			return antigravity429RateLimited
+		}
+	}
+	return antigravity429Unknown
+}
+
+func antigravityCreditsRetryEnabled(cfg *config.Config) bool {
+	return cfg != nil && cfg.QuotaExceeded.AntigravityCredits
+}
+
+func antigravityCreditsExhausted(auth *cliproxyauth.Auth, now time.Time) bool {
+	if auth == nil || strings.TrimSpace(auth.ID) == "" {
+		return false
+	}
+	value, ok := antigravityCreditsExhaustedByAuth.Load(auth.ID)
+	if !ok {
+		return false
+	}
+	until, ok := value.(time.Time)
+	if !ok || until.IsZero() {
+		antigravityCreditsExhaustedByAuth.Delete(auth.ID)
+		return false
+	}
+	if !until.After(now) {
+		antigravityCreditsExhaustedByAuth.Delete(auth.ID)
+		return false
+	}
+	return true
+}
+
+func markAntigravityCreditsExhausted(auth *cliproxyauth.Auth, now time.Time) {
+	if auth == nil || strings.TrimSpace(auth.ID) == "" {
+		return
+	}
+	antigravityCreditsExhaustedByAuth.Store(auth.ID, now.Add(antigravityCreditsRetryTTL))
+}
+
+func clearAntigravityCreditsExhausted(auth *cliproxyauth.Auth) {
+	if auth == nil || strings.TrimSpace(auth.ID) == "" {
+		return
+	}
+	antigravityCreditsExhaustedByAuth.Delete(auth.ID)
+}
+
+func antigravityPreferCreditsKey(auth *cliproxyauth.Auth, modelName string) string {
+	if auth == nil {
+		return ""
+	}
+	authID := strings.TrimSpace(auth.ID)
+	modelName = strings.TrimSpace(modelName)
+	if authID == "" || modelName == "" {
+		return ""
+	}
+	return authID + "|" + modelName
+}
+
+func antigravityShouldPreferCredits(auth *cliproxyauth.Auth, modelName string, now time.Time) bool {
+	key := antigravityPreferCreditsKey(auth, modelName)
+	if key == "" {
+		return false
+	}
+	value, ok := antigravityPreferCreditsByModel.Load(key)
+	if !ok {
+		return false
+	}
+	until, ok := value.(time.Time)
+	if !ok || until.IsZero() {
+		antigravityPreferCreditsByModel.Delete(key)
+		return false
+	}
+	if !until.After(now) {
+		antigravityPreferCreditsByModel.Delete(key)
+		return false
+	}
+	return true
+}
+
+func markAntigravityPreferCredits(auth *cliproxyauth.Auth, modelName string, now time.Time, retryAfter *time.Duration) {
+	key := antigravityPreferCreditsKey(auth, modelName)
+	if key == "" {
+		return
+	}
+	until := now.Add(antigravityCreditsRetryTTL)
+	if retryAfter != nil && *retryAfter > 0 {
+		until = now.Add(*retryAfter)
+	}
+	antigravityPreferCreditsByModel.Store(key, until)
+}
+
+func clearAntigravityPreferCredits(auth *cliproxyauth.Auth, modelName string) {
+	key := antigravityPreferCreditsKey(auth, modelName)
+	if key == "" {
+		return
+	}
+	antigravityPreferCreditsByModel.Delete(key)
+}
+
+func shouldMarkAntigravityCreditsExhausted(statusCode int, body []byte, reqErr error) bool {
+	if reqErr != nil || statusCode == 0 {
+		return false
+	}
+	if statusCode >= http.StatusInternalServerError || statusCode == http.StatusRequestTimeout {
+		return false
+	}
+	lowerBody := strings.ToLower(string(body))
+	for _, keyword := range antigravityCreditsExhaustedKeywords {
+		if strings.Contains(lowerBody, keyword) {
+			return true
+		}
+	}
+	return false
+}
+
+func newAntigravityStatusErr(statusCode int, body []byte) statusErr {
+	err := statusErr{code: statusCode, msg: string(body)}
+	if statusCode == http.StatusTooManyRequests {
+		if retryAfter, parseErr := parseRetryDelay(body); parseErr == nil && retryAfter != nil {
+			err.retryAfter = retryAfter
+		}
+	}
+	return err
+}
+
+func (e *AntigravityExecutor) attemptCreditsFallback(
+	ctx context.Context,
+	auth *cliproxyauth.Auth,
+	httpClient *http.Client,
+	token string,
+	modelName string,
+	payload []byte,
+	stream bool,
+	alt string,
+	baseURL string,
+	originalBody []byte,
+) (*http.Response, bool) {
+	if !antigravityCreditsRetryEnabled(e.cfg) {
+		return nil, false
+	}
+	if classifyAntigravity429(originalBody) != antigravity429QuotaExhausted {
+		return nil, false
+	}
+	now := time.Now()
+	if antigravityCreditsExhausted(auth, now) {
+		return nil, false
+	}
+	creditsPayload := injectEnabledCreditTypes(payload)
+	if len(creditsPayload) == 0 {
+		return nil, false
+	}
+
+	httpReq, errReq := e.buildRequest(ctx, auth, token, modelName, creditsPayload, stream, alt, baseURL)
+	if errReq != nil {
+		helps.RecordAPIResponseError(ctx, e.cfg, errReq)
+		return nil, true
+	}
+	httpResp, errDo := httpClient.Do(httpReq)
+	if errDo != nil {
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
+		return nil, true
+	}
+	if httpResp.StatusCode >= http.StatusOK && httpResp.StatusCode < http.StatusMultipleChoices {
+		retryAfter, _ := parseRetryDelay(originalBody)
+		markAntigravityPreferCredits(auth, modelName, now, retryAfter)
+		clearAntigravityCreditsExhausted(auth)
+		return httpResp, true
+	}
+
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	bodyBytes, errRead := io.ReadAll(httpResp.Body)
+	if errClose := httpResp.Body.Close(); errClose != nil {
+		log.Errorf("antigravity executor: close credits fallback response body error: %v", errClose)
+	}
+	if errRead != nil {
+		helps.RecordAPIResponseError(ctx, e.cfg, errRead)
+		return nil, true
+	}
+	helps.AppendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+	if shouldMarkAntigravityCreditsExhausted(httpResp.StatusCode, bodyBytes, nil) {
+		clearAntigravityPreferCredits(auth, modelName)
+		markAntigravityCreditsExhausted(auth, now)
+	}
+	return nil, true
+}
+
 // Execute performs a non-streaming request to the Antigravity API.
 func (e *AntigravityExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options) (resp cliproxyexecutor.Response, err error) {
 	if opts.Alt == "responses/compact" {
@@ -203,8 +458,8 @@ func (e *AntigravityExecutor) Execute(ctx context.Context, auth *cliproxyauth.Au
 		auth = updatedAuth
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("antigravity")
@@ -222,8 +477,8 @@ func (e *AntigravityExecutor) Execute(ctx context.Context, auth *cliproxyauth.Au
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	translated = applyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	translated = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
 
 	baseURLs := antigravityBaseURLFallbackOrder(auth)
 	httpClient := newAntigravityHTTPClient(ctx, e.cfg, auth, 0)
@@ -237,7 +492,15 @@ attemptLoop:
 		var lastErr error
 
 		for idx, baseURL := range baseURLs {
-			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, translated, false, opts.Alt, baseURL)
+			requestPayload := translated
+			usedCreditsDirect := false
+			if antigravityCreditsRetryEnabled(e.cfg) && antigravityShouldPreferCredits(auth, baseModel, time.Now()) {
+				if creditsPayload := injectEnabledCreditTypes(translated); len(creditsPayload) > 0 {
+					requestPayload = creditsPayload
+					usedCreditsDirect = true
+				}
+			}
+			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, requestPayload, false, opts.Alt, baseURL)
 			if errReq != nil {
 				err = errReq
 				return resp, err
@@ -245,7 +508,7 @@ attemptLoop:
 
 			httpResp, errDo := httpClient.Do(httpReq)
 			if errDo != nil {
-				recordAPIResponseError(ctx, e.cfg, errDo)
+				helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 				if errors.Is(errDo, context.Canceled) || errors.Is(errDo, context.DeadlineExceeded) {
 					return resp, errDo
 				}
@@ -260,20 +523,50 @@ attemptLoop:
 				return resp, err
 			}
 
-			recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+			helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 			bodyBytes, errRead := io.ReadAll(httpResp.Body)
 			if errClose := httpResp.Body.Close(); errClose != nil {
 				log.Errorf("antigravity executor: close response body error: %v", errClose)
 			}
 			if errRead != nil {
-				recordAPIResponseError(ctx, e.cfg, errRead)
+				helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 				err = errRead
 				return resp, err
 			}
-			appendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+
+			if httpResp.StatusCode == http.StatusTooManyRequests {
+				if usedCreditsDirect {
+					if shouldMarkAntigravityCreditsExhausted(httpResp.StatusCode, bodyBytes, nil) {
+						clearAntigravityPreferCredits(auth, baseModel)
+						markAntigravityCreditsExhausted(auth, time.Now())
+					}
+				} else {
+					creditsResp, _ := e.attemptCreditsFallback(ctx, auth, httpClient, token, baseModel, translated, false, opts.Alt, baseURL, bodyBytes)
+					if creditsResp != nil {
+						helps.RecordAPIResponseMetadata(ctx, e.cfg, creditsResp.StatusCode, creditsResp.Header.Clone())
+						creditsBody, errCreditsRead := io.ReadAll(creditsResp.Body)
+						if errClose := creditsResp.Body.Close(); errClose != nil {
+							log.Errorf("antigravity executor: close credits success response body error: %v", errClose)
+						}
+						if errCreditsRead != nil {
+							helps.RecordAPIResponseError(ctx, e.cfg, errCreditsRead)
+							err = errCreditsRead
+							return resp, err
+						}
+						helps.AppendAPIResponseChunk(ctx, e.cfg, creditsBody)
+						reporter.Publish(ctx, helps.ParseAntigravityUsage(creditsBody))
+						var param any
+						converted := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, translated, creditsBody, &param)
+						resp = cliproxyexecutor.Response{Payload: converted, Headers: creditsResp.Header.Clone()}
+						reporter.EnsurePublished(ctx)
+						return resp, nil
+					}
+				}
+			}
 
 			if httpResp.StatusCode < http.StatusOK || httpResp.StatusCode >= http.StatusMultipleChoices {
-				log.Debugf("antigravity executor: upstream error status: %d, body: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), bodyBytes))
+				log.Debugf("antigravity executor: upstream error status: %d, body: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), bodyBytes))
 				lastStatus = httpResp.StatusCode
 				lastBody = append([]byte(nil), bodyBytes...)
 				lastErr = nil
@@ -295,33 +588,21 @@ attemptLoop:
 						continue attemptLoop
 					}
 				}
-				sErr := statusErr{code: httpResp.StatusCode, msg: string(bodyBytes)}
-				if httpResp.StatusCode == http.StatusTooManyRequests {
-					if retryAfter, parseErr := parseRetryDelay(bodyBytes); parseErr == nil && retryAfter != nil {
-						sErr.retryAfter = retryAfter
-					}
-				}
-				err = sErr
+				err = newAntigravityStatusErr(httpResp.StatusCode, bodyBytes)
 				return resp, err
 			}
 
-			reporter.publish(ctx, parseAntigravityUsage(bodyBytes))
+			reporter.Publish(ctx, helps.ParseAntigravityUsage(bodyBytes))
 			var param any
 			converted := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, translated, bodyBytes, &param)
 			resp = cliproxyexecutor.Response{Payload: converted, Headers: httpResp.Header.Clone()}
-			reporter.ensurePublished(ctx)
+			reporter.EnsurePublished(ctx)
 			return resp, nil
 		}
 
 		switch {
 		case lastStatus != 0:
-			sErr := statusErr{code: lastStatus, msg: string(lastBody)}
-			if lastStatus == http.StatusTooManyRequests {
-				if retryAfter, parseErr := parseRetryDelay(lastBody); parseErr == nil && retryAfter != nil {
-					sErr.retryAfter = retryAfter
-				}
-			}
-			err = sErr
+			err = newAntigravityStatusErr(lastStatus, lastBody)
 		case lastErr != nil:
 			err = lastErr
 		default:
@@ -345,8 +626,8 @@ func (e *AntigravityExecutor) executeClaudeNonStream(ctx context.Context, auth *
 		auth = updatedAuth
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("antigravity")
@@ -364,8 +645,8 @@ func (e *AntigravityExecutor) executeClaudeNonStream(ctx context.Context, auth *
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	translated = applyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	translated = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
 
 	baseURLs := antigravityBaseURLFallbackOrder(auth)
 	httpClient := newAntigravityHTTPClient(ctx, e.cfg, auth, 0)
@@ -379,7 +660,15 @@ attemptLoop:
 		var lastErr error
 
 		for idx, baseURL := range baseURLs {
-			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, translated, true, opts.Alt, baseURL)
+			requestPayload := translated
+			usedCreditsDirect := false
+			if antigravityCreditsRetryEnabled(e.cfg) && antigravityShouldPreferCredits(auth, baseModel, time.Now()) {
+				if creditsPayload := injectEnabledCreditTypes(translated); len(creditsPayload) > 0 {
+					requestPayload = creditsPayload
+					usedCreditsDirect = true
+				}
+			}
+			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, requestPayload, true, opts.Alt, baseURL)
 			if errReq != nil {
 				err = errReq
 				return resp, err
@@ -387,7 +676,7 @@ attemptLoop:
 
 			httpResp, errDo := httpClient.Do(httpReq)
 			if errDo != nil {
-				recordAPIResponseError(ctx, e.cfg, errDo)
+				helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 				if errors.Is(errDo, context.Canceled) || errors.Is(errDo, context.DeadlineExceeded) {
 					return resp, errDo
 				}
@@ -401,14 +690,14 @@ attemptLoop:
 				err = errDo
 				return resp, err
 			}
-			recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+			helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 			if httpResp.StatusCode < http.StatusOK || httpResp.StatusCode >= http.StatusMultipleChoices {
 				bodyBytes, errRead := io.ReadAll(httpResp.Body)
 				if errClose := httpResp.Body.Close(); errClose != nil {
 					log.Errorf("antigravity executor: close response body error: %v", errClose)
 				}
 				if errRead != nil {
-					recordAPIResponseError(ctx, e.cfg, errRead)
+					helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 					if errors.Is(errRead, context.Canceled) || errors.Is(errRead, context.DeadlineExceeded) {
 						err = errRead
 						return resp, err
@@ -427,7 +716,24 @@ attemptLoop:
 					err = errRead
 					return resp, err
 				}
-				appendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+				helps.AppendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+				if httpResp.StatusCode == http.StatusTooManyRequests {
+					if usedCreditsDirect {
+						if shouldMarkAntigravityCreditsExhausted(httpResp.StatusCode, bodyBytes, nil) {
+							clearAntigravityPreferCredits(auth, baseModel)
+							markAntigravityCreditsExhausted(auth, time.Now())
+						}
+					} else {
+						creditsResp, _ := e.attemptCreditsFallback(ctx, auth, httpClient, token, baseModel, translated, true, opts.Alt, baseURL, bodyBytes)
+						if creditsResp != nil {
+							httpResp = creditsResp
+							helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+						}
+					}
+				}
+				if httpResp.StatusCode >= http.StatusOK && httpResp.StatusCode < http.StatusMultipleChoices {
+					goto streamSuccessClaudeNonStream
+				}
 				lastStatus = httpResp.StatusCode
 				lastBody = append([]byte(nil), bodyBytes...)
 				lastErr = nil
@@ -449,16 +755,11 @@ attemptLoop:
 						continue attemptLoop
 					}
 				}
-				sErr := statusErr{code: httpResp.StatusCode, msg: string(bodyBytes)}
-				if httpResp.StatusCode == http.StatusTooManyRequests {
-					if retryAfter, parseErr := parseRetryDelay(bodyBytes); parseErr == nil && retryAfter != nil {
-						sErr.retryAfter = retryAfter
-					}
-				}
-				err = sErr
+				err = newAntigravityStatusErr(httpResp.StatusCode, bodyBytes)
 				return resp, err
 			}
 
+		streamSuccessClaudeNonStream:
 			out := make(chan cliproxyexecutor.StreamChunk)
 			go func(resp *http.Response) {
 				defer close(out)
@@ -471,29 +772,29 @@ attemptLoop:
 				scanner.Buffer(nil, streamScannerBuffer)
 				for scanner.Scan() {
 					line := scanner.Bytes()
-					appendAPIResponseChunk(ctx, e.cfg, line)
+					helps.AppendAPIResponseChunk(ctx, e.cfg, line)
 
 					// Filter usage metadata for all models
 					// Only retain usage statistics in the terminal chunk
-					line = FilterSSEUsageMetadata(line)
+					line = helps.FilterSSEUsageMetadata(line)
 
-					payload := jsonPayload(line)
+					payload := helps.JSONPayload(line)
 					if payload == nil {
 						continue
 					}
 
-					if detail, ok := parseAntigravityStreamUsage(payload); ok {
-						reporter.publish(ctx, detail)
+					if detail, ok := helps.ParseAntigravityStreamUsage(payload); ok {
+						reporter.Publish(ctx, detail)
 					}
 
 					out <- cliproxyexecutor.StreamChunk{Payload: payload}
 				}
 				if errScan := scanner.Err(); errScan != nil {
-					recordAPIResponseError(ctx, e.cfg, errScan)
-					reporter.publishFailure(ctx)
+					helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+					reporter.PublishFailure(ctx)
 					out <- cliproxyexecutor.StreamChunk{Err: errScan}
 				} else {
-					reporter.ensurePublished(ctx)
+					reporter.EnsurePublished(ctx)
 				}
 			}(httpResp)
 
@@ -509,24 +810,18 @@ attemptLoop:
 			}
 			resp = cliproxyexecutor.Response{Payload: e.convertStreamToNonStream(buffer.Bytes())}
 
-			reporter.publish(ctx, parseAntigravityUsage(resp.Payload))
+			reporter.Publish(ctx, helps.ParseAntigravityUsage(resp.Payload))
 			var param any
 			converted := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, translated, resp.Payload, &param)
 			resp = cliproxyexecutor.Response{Payload: converted, Headers: httpResp.Header.Clone()}
-			reporter.ensurePublished(ctx)
+			reporter.EnsurePublished(ctx)
 
 			return resp, nil
 		}
 
 		switch {
 		case lastStatus != 0:
-			sErr := statusErr{code: lastStatus, msg: string(lastBody)}
-			if lastStatus == http.StatusTooManyRequests {
-				if retryAfter, parseErr := parseRetryDelay(lastBody); parseErr == nil && retryAfter != nil {
-					sErr.retryAfter = retryAfter
-				}
-			}
-			err = sErr
+			err = newAntigravityStatusErr(lastStatus, lastBody)
 		case lastErr != nil:
 			err = lastErr
 		default:
@@ -748,8 +1043,8 @@ func (e *AntigravityExecutor) ExecuteStream(ctx context.Context, auth *cliproxya
 		auth = updatedAuth
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("antigravity")
@@ -767,8 +1062,8 @@ func (e *AntigravityExecutor) ExecuteStream(ctx context.Context, auth *cliproxya
 		return nil, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	translated = applyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	translated = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, "antigravity", "request", translated, originalTranslated, requestedModel)
 
 	baseURLs := antigravityBaseURLFallbackOrder(auth)
 	httpClient := newAntigravityHTTPClient(ctx, e.cfg, auth, 0)
@@ -782,14 +1077,22 @@ attemptLoop:
 		var lastErr error
 
 		for idx, baseURL := range baseURLs {
-			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, translated, true, opts.Alt, baseURL)
+			requestPayload := translated
+			usedCreditsDirect := false
+			if antigravityCreditsRetryEnabled(e.cfg) && antigravityShouldPreferCredits(auth, baseModel, time.Now()) {
+				if creditsPayload := injectEnabledCreditTypes(translated); len(creditsPayload) > 0 {
+					requestPayload = creditsPayload
+					usedCreditsDirect = true
+				}
+			}
+			httpReq, errReq := e.buildRequest(ctx, auth, token, baseModel, requestPayload, true, opts.Alt, baseURL)
 			if errReq != nil {
 				err = errReq
 				return nil, err
 			}
 			httpResp, errDo := httpClient.Do(httpReq)
 			if errDo != nil {
-				recordAPIResponseError(ctx, e.cfg, errDo)
+				helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 				if errors.Is(errDo, context.Canceled) || errors.Is(errDo, context.DeadlineExceeded) {
 					return nil, errDo
 				}
@@ -803,14 +1106,14 @@ attemptLoop:
 				err = errDo
 				return nil, err
 			}
-			recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+			helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 			if httpResp.StatusCode < http.StatusOK || httpResp.StatusCode >= http.StatusMultipleChoices {
 				bodyBytes, errRead := io.ReadAll(httpResp.Body)
 				if errClose := httpResp.Body.Close(); errClose != nil {
 					log.Errorf("antigravity executor: close response body error: %v", errClose)
 				}
 				if errRead != nil {
-					recordAPIResponseError(ctx, e.cfg, errRead)
+					helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 					if errors.Is(errRead, context.Canceled) || errors.Is(errRead, context.DeadlineExceeded) {
 						err = errRead
 						return nil, err
@@ -829,7 +1132,24 @@ attemptLoop:
 					err = errRead
 					return nil, err
 				}
-				appendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+				helps.AppendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+				if httpResp.StatusCode == http.StatusTooManyRequests {
+					if usedCreditsDirect {
+						if shouldMarkAntigravityCreditsExhausted(httpResp.StatusCode, bodyBytes, nil) {
+							clearAntigravityPreferCredits(auth, baseModel)
+							markAntigravityCreditsExhausted(auth, time.Now())
+						}
+					} else {
+						creditsResp, _ := e.attemptCreditsFallback(ctx, auth, httpClient, token, baseModel, translated, true, opts.Alt, baseURL, bodyBytes)
+						if creditsResp != nil {
+							httpResp = creditsResp
+							helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+						}
+					}
+				}
+				if httpResp.StatusCode >= http.StatusOK && httpResp.StatusCode < http.StatusMultipleChoices {
+					goto streamSuccessExecuteStream
+				}
 				lastStatus = httpResp.StatusCode
 				lastBody = append([]byte(nil), bodyBytes...)
 				lastErr = nil
@@ -851,16 +1171,11 @@ attemptLoop:
 						continue attemptLoop
 					}
 				}
-				sErr := statusErr{code: httpResp.StatusCode, msg: string(bodyBytes)}
-				if httpResp.StatusCode == http.StatusTooManyRequests {
-					if retryAfter, parseErr := parseRetryDelay(bodyBytes); parseErr == nil && retryAfter != nil {
-						sErr.retryAfter = retryAfter
-					}
-				}
-				err = sErr
+				err = newAntigravityStatusErr(httpResp.StatusCode, bodyBytes)
 				return nil, err
 			}
 
+		streamSuccessExecuteStream:
 			out := make(chan cliproxyexecutor.StreamChunk)
 			go func(resp *http.Response) {
 				defer close(out)
@@ -874,19 +1189,19 @@ attemptLoop:
 				var param any
 				for scanner.Scan() {
 					line := scanner.Bytes()
-					appendAPIResponseChunk(ctx, e.cfg, line)
+					helps.AppendAPIResponseChunk(ctx, e.cfg, line)
 
 					// Filter usage metadata for all models
 					// Only retain usage statistics in the terminal chunk
-					line = FilterSSEUsageMetadata(line)
+					line = helps.FilterSSEUsageMetadata(line)
 
-					payload := jsonPayload(line)
+					payload := helps.JSONPayload(line)
 					if payload == nil {
 						continue
 					}
 
-					if detail, ok := parseAntigravityStreamUsage(payload); ok {
-						reporter.publish(ctx, detail)
+					if detail, ok := helps.ParseAntigravityStreamUsage(payload); ok {
+						reporter.Publish(ctx, detail)
 					}
 
 					chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, translated, bytes.Clone(payload), &param)
@@ -899,11 +1214,11 @@ attemptLoop:
 					out <- cliproxyexecutor.StreamChunk{Payload: tail[i]}
 				}
 				if errScan := scanner.Err(); errScan != nil {
-					recordAPIResponseError(ctx, e.cfg, errScan)
-					reporter.publishFailure(ctx)
+					helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+					reporter.PublishFailure(ctx)
 					out <- cliproxyexecutor.StreamChunk{Err: errScan}
 				} else {
-					reporter.ensurePublished(ctx)
+					reporter.EnsurePublished(ctx)
 				}
 			}(httpResp)
 			return &cliproxyexecutor.StreamResult{Headers: httpResp.Header.Clone(), Chunks: out}, nil
@@ -911,13 +1226,7 @@ attemptLoop:
 
 		switch {
 		case lastStatus != 0:
-			sErr := statusErr{code: lastStatus, msg: string(lastBody)}
-			if lastStatus == http.StatusTooManyRequests {
-				if retryAfter, parseErr := parseRetryDelay(lastBody); parseErr == nil && retryAfter != nil {
-					sErr.retryAfter = retryAfter
-				}
-			}
-			err = sErr
+			err = newAntigravityStatusErr(lastStatus, lastBody)
 		case lastErr != nil:
 			err = lastErr
 		default:
@@ -1011,8 +1320,13 @@ func (e *AntigravityExecutor) CountTokens(ctx context.Context, auth *cliproxyaut
 		if host := resolveHost(base); host != "" {
 			httpReq.Host = host
 		}
+		var attrs map[string]string
+		if auth != nil {
+			attrs = auth.Attributes
+		}
+		util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
-		recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+		helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 			URL:       requestURL.String(),
 			Method:    http.MethodPost,
 			Headers:   httpReq.Header.Clone(),
@@ -1026,7 +1340,7 @@ func (e *AntigravityExecutor) CountTokens(ctx context.Context, auth *cliproxyaut
 
 		httpResp, errDo := httpClient.Do(httpReq)
 		if errDo != nil {
-			recordAPIResponseError(ctx, e.cfg, errDo)
+			helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 			if errors.Is(errDo, context.Canceled) || errors.Is(errDo, context.DeadlineExceeded) {
 				return cliproxyexecutor.Response{}, errDo
 			}
@@ -1040,16 +1354,16 @@ func (e *AntigravityExecutor) CountTokens(ctx context.Context, auth *cliproxyaut
 			return cliproxyexecutor.Response{}, errDo
 		}
 
-		recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+		helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 		bodyBytes, errRead := io.ReadAll(httpResp.Body)
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("antigravity executor: close response body error: %v", errClose)
 		}
 		if errRead != nil {
-			recordAPIResponseError(ctx, e.cfg, errRead)
+			helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 			return cliproxyexecutor.Response{}, errRead
 		}
-		appendAPIResponseChunk(ctx, e.cfg, bodyBytes)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, bodyBytes)
 
 		if httpResp.StatusCode >= http.StatusOK && httpResp.StatusCode < http.StatusMultipleChoices {
 			count := gjson.GetBytes(bodyBytes, "totalTokens").Int()
@@ -1305,6 +1619,11 @@ func (e *AntigravityExecutor) buildRequest(ctx context.Context, auth *cliproxyau
 	if host := resolveHost(base); host != "" {
 		httpReq.Host = host
 	}
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -1316,7 +1635,7 @@ func (e *AntigravityExecutor) buildRequest(ctx context.Context, auth *cliproxyau
 	if e.cfg != nil && e.cfg.RequestLog {
 		payloadLog = []byte(payloadStr)
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       requestURL.String(),
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -1479,7 +1798,7 @@ func antigravityWait(ctx context.Context, wait time.Duration) error {
 	}
 }
 
-func antigravityBaseURLFallbackOrder(auth *cliproxyauth.Auth) []string {
+var antigravityBaseURLFallbackOrder = func(auth *cliproxyauth.Auth) []string {
 	if base := resolveCustomAntigravityBaseURL(auth); base != "" {
 		return []string{base}
 	}

internal/runtime/executor/antigravity_executor_credits_test.go

@@ -0,0 +1,423 @@
+package executor
+
+import (
+	"context"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
+	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
+)
+
+func resetAntigravityCreditsRetryState() {
+	antigravityCreditsExhaustedByAuth = sync.Map{}
+	antigravityPreferCreditsByModel = sync.Map{}
+}
+
+func TestClassifyAntigravity429(t *testing.T) {
+	t.Run("quota exhausted", func(t *testing.T) {
+		body := []byte(`{"error":{"status":"RESOURCE_EXHAUSTED","message":"QUOTA_EXHAUSTED"}}`)
+		if got := classifyAntigravity429(body); got != antigravity429QuotaExhausted {
+			t.Fatalf("classifyAntigravity429() = %q, want %q", got, antigravity429QuotaExhausted)
+		}
+	})
+
+	t.Run("structured rate limit", func(t *testing.T) {
+		body := []byte(`{
+			"error": {
+				"status": "RESOURCE_EXHAUSTED",
+				"details": [
+					{"@type": "type.googleapis.com/google.rpc.ErrorInfo", "reason": "RATE_LIMIT_EXCEEDED"},
+					{"@type": "type.googleapis.com/google.rpc.RetryInfo", "retryDelay": "0.5s"}
+				]
+			}
+		}`)
+		if got := classifyAntigravity429(body); got != antigravity429RateLimited {
+			t.Fatalf("classifyAntigravity429() = %q, want %q", got, antigravity429RateLimited)
+		}
+	})
+
+	t.Run("structured quota exhausted", func(t *testing.T) {
+		body := []byte(`{
+			"error": {
+				"status": "RESOURCE_EXHAUSTED",
+				"details": [
+					{"@type": "type.googleapis.com/google.rpc.ErrorInfo", "reason": "QUOTA_EXHAUSTED"}
+				]
+			}
+		}`)
+		if got := classifyAntigravity429(body); got != antigravity429QuotaExhausted {
+			t.Fatalf("classifyAntigravity429() = %q, want %q", got, antigravity429QuotaExhausted)
+		}
+	})
+
+	t.Run("unknown", func(t *testing.T) {
+		body := []byte(`{"error":{"message":"too many requests"}}`)
+		if got := classifyAntigravity429(body); got != antigravity429Unknown {
+			t.Fatalf("classifyAntigravity429() = %q, want %q", got, antigravity429Unknown)
+		}
+	})
+}
+
+func TestInjectEnabledCreditTypes(t *testing.T) {
+	body := []byte(`{"model":"gemini-2.5-flash","request":{}}`)
+	got := injectEnabledCreditTypes(body)
+	if got == nil {
+		t.Fatal("injectEnabledCreditTypes() returned nil")
+	}
+	if !strings.Contains(string(got), `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+		t.Fatalf("injectEnabledCreditTypes() = %s, want enabledCreditTypes", string(got))
+	}
+
+	if got := injectEnabledCreditTypes([]byte(`not json`)); got != nil {
+		t.Fatalf("injectEnabledCreditTypes() for invalid json = %s, want nil", string(got))
+	}
+}
+
+func TestShouldMarkAntigravityCreditsExhausted(t *testing.T) {
+	for _, body := range [][]byte{
+		[]byte(`{"error":{"message":"Insufficient GOOGLE_ONE_AI credits"}}`),
+		[]byte(`{"error":{"message":"minimumCreditAmountForUsage requirement not met"}}`),
+		[]byte(`{"error":{"message":"Resource has been exhausted"}}`),
+	} {
+		if !shouldMarkAntigravityCreditsExhausted(http.StatusForbidden, body, nil) {
+			t.Fatalf("shouldMarkAntigravityCreditsExhausted(%s) = false, want true", string(body))
+		}
+	}
+	if shouldMarkAntigravityCreditsExhausted(http.StatusServiceUnavailable, []byte(`{"error":{"message":"credits exhausted"}}`), nil) {
+		t.Fatal("shouldMarkAntigravityCreditsExhausted() = true for 5xx, want false")
+	}
+}
+
+func TestAntigravityExecute_RetriesQuotaExhaustedWithCredits(t *testing.T) {
+	resetAntigravityCreditsRetryState()
+	t.Cleanup(resetAntigravityCreditsRetryState)
+
+	var (
+		mu            sync.Mutex
+		requestBodies []string
+	)
+
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		_ = r.Body.Close()
+
+		mu.Lock()
+		requestBodies = append(requestBodies, string(body))
+		reqNum := len(requestBodies)
+		mu.Unlock()
+
+		if reqNum == 1 {
+			w.WriteHeader(http.StatusTooManyRequests)
+			_, _ = w.Write([]byte(`{"error":{"status":"RESOURCE_EXHAUSTED","message":"QUOTA_EXHAUSTED"}}`))
+			return
+		}
+
+		if !strings.Contains(string(body), `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+			t.Fatalf("second request body missing enabledCreditTypes: %s", string(body))
+		}
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"response":{"candidates":[{"content":{"role":"model","parts":[{"text":"ok"}]}}],"usageMetadata":{"promptTokenCount":1,"candidatesTokenCount":1,"totalTokenCount":2}}}`))
+	}))
+	defer server.Close()
+
+	exec := NewAntigravityExecutor(&config.Config{
+		QuotaExceeded: config.QuotaExceeded{AntigravityCredits: true},
+	})
+	auth := &cliproxyauth.Auth{
+		ID: "auth-credits-ok",
+		Attributes: map[string]string{
+			"base_url": server.URL,
+		},
+		Metadata: map[string]any{
+			"access_token": "token",
+			"project_id":   "project-1",
+			"expired":      time.Now().Add(1 * time.Hour).Format(time.RFC3339),
+		},
+	}
+
+	resp, err := exec.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gemini-2.5-flash",
+		Payload: []byte(`{"request":{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FormatAntigravity,
+	})
+	if err != nil {
+		t.Fatalf("Execute() error = %v", err)
+	}
+	if len(resp.Payload) == 0 {
+		t.Fatal("Execute() returned empty payload")
+	}
+
+	mu.Lock()
+	defer mu.Unlock()
+	if len(requestBodies) != 2 {
+		t.Fatalf("request count = %d, want 2", len(requestBodies))
+	}
+}
+
+func TestAntigravityExecute_SkipsCreditsRetryWhenAlreadyExhausted(t *testing.T) {
+	resetAntigravityCreditsRetryState()
+	t.Cleanup(resetAntigravityCreditsRetryState)
+
+	var requestCount int
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		requestCount++
+		w.WriteHeader(http.StatusTooManyRequests)
+		_, _ = w.Write([]byte(`{"error":{"status":"RESOURCE_EXHAUSTED","message":"QUOTA_EXHAUSTED"}}`))
+	}))
+	defer server.Close()
+
+	exec := NewAntigravityExecutor(&config.Config{
+		QuotaExceeded: config.QuotaExceeded{AntigravityCredits: true},
+	})
+	auth := &cliproxyauth.Auth{
+		ID: "auth-credits-exhausted",
+		Attributes: map[string]string{
+			"base_url": server.URL,
+		},
+		Metadata: map[string]any{
+			"access_token": "token",
+			"project_id":   "project-1",
+			"expired":      time.Now().Add(1 * time.Hour).Format(time.RFC3339),
+		},
+	}
+	markAntigravityCreditsExhausted(auth, time.Now())
+
+	_, err := exec.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gemini-2.5-flash",
+		Payload: []byte(`{"request":{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FormatAntigravity,
+	})
+	if err == nil {
+		t.Fatal("Execute() error = nil, want 429")
+	}
+	sErr, ok := err.(statusErr)
+	if !ok {
+		t.Fatalf("Execute() error type = %T, want statusErr", err)
+	}
+	if got := sErr.StatusCode(); got != http.StatusTooManyRequests {
+		t.Fatalf("Execute() status code = %d, want %d", got, http.StatusTooManyRequests)
+	}
+	if requestCount != 1 {
+		t.Fatalf("request count = %d, want 1", requestCount)
+	}
+}
+
+func TestAntigravityExecute_PrefersCreditsAfterSuccessfulFallback(t *testing.T) {
+	resetAntigravityCreditsRetryState()
+	t.Cleanup(resetAntigravityCreditsRetryState)
+
+	var (
+		mu            sync.Mutex
+		requestBodies []string
+	)
+
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		_ = r.Body.Close()
+
+		mu.Lock()
+		requestBodies = append(requestBodies, string(body))
+		reqNum := len(requestBodies)
+		mu.Unlock()
+
+		switch reqNum {
+		case 1:
+			w.WriteHeader(http.StatusTooManyRequests)
+			_, _ = w.Write([]byte(`{"error":{"status":"RESOURCE_EXHAUSTED","details":[{"@type":"type.googleapis.com/google.rpc.ErrorInfo","reason":"QUOTA_EXHAUSTED"},{"@type":"type.googleapis.com/google.rpc.RetryInfo","retryDelay":"10s"}]}}`))
+		case 2, 3:
+			if !strings.Contains(string(body), `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+				t.Fatalf("request %d body missing enabledCreditTypes: %s", reqNum, string(body))
+			}
+			w.Header().Set("Content-Type", "application/json")
+			_, _ = w.Write([]byte(`{"response":{"candidates":[{"content":{"role":"model","parts":[{"text":"OK"}]}}],"usageMetadata":{"promptTokenCount":1,"candidatesTokenCount":1,"totalTokenCount":2}}}`))
+		default:
+			t.Fatalf("unexpected request count %d", reqNum)
+		}
+	}))
+	defer server.Close()
+
+	exec := NewAntigravityExecutor(&config.Config{
+		QuotaExceeded: config.QuotaExceeded{AntigravityCredits: true},
+	})
+	auth := &cliproxyauth.Auth{
+		ID: "auth-prefer-credits",
+		Attributes: map[string]string{
+			"base_url": server.URL,
+		},
+		Metadata: map[string]any{
+			"access_token": "token",
+			"project_id":   "project-1",
+			"expired":      time.Now().Add(1 * time.Hour).Format(time.RFC3339),
+		},
+	}
+
+	request := cliproxyexecutor.Request{
+		Model:   "gemini-2.5-flash",
+		Payload: []byte(`{"request":{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}}`),
+	}
+	opts := cliproxyexecutor.Options{SourceFormat: sdktranslator.FormatAntigravity}
+
+	if _, err := exec.Execute(context.Background(), auth, request, opts); err != nil {
+		t.Fatalf("first Execute() error = %v", err)
+	}
+	if _, err := exec.Execute(context.Background(), auth, request, opts); err != nil {
+		t.Fatalf("second Execute() error = %v", err)
+	}
+
+	mu.Lock()
+	defer mu.Unlock()
+	if len(requestBodies) != 3 {
+		t.Fatalf("request count = %d, want 3", len(requestBodies))
+	}
+	if strings.Contains(requestBodies[0], `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+		t.Fatalf("first request unexpectedly used credits: %s", requestBodies[0])
+	}
+	if !strings.Contains(requestBodies[1], `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+		t.Fatalf("fallback request missing credits: %s", requestBodies[1])
+	}
+	if !strings.Contains(requestBodies[2], `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+		t.Fatalf("preferred request missing credits: %s", requestBodies[2])
+	}
+}
+
+func TestAntigravityExecute_PreservesBaseURLFallbackAfterCreditsRetryFailure(t *testing.T) {
+	resetAntigravityCreditsRetryState()
+	t.Cleanup(resetAntigravityCreditsRetryState)
+
+	var (
+		mu          sync.Mutex
+		firstCount  int
+		secondCount int
+	)
+
+	firstServer := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		_ = r.Body.Close()
+
+		mu.Lock()
+		firstCount++
+		reqNum := firstCount
+		mu.Unlock()
+
+		switch reqNum {
+		case 1:
+			w.WriteHeader(http.StatusTooManyRequests)
+			_, _ = w.Write([]byte(`{"error":{"status":"RESOURCE_EXHAUSTED","details":[{"@type":"type.googleapis.com/google.rpc.ErrorInfo","reason":"QUOTA_EXHAUSTED"}]}}`))
+		case 2:
+			if !strings.Contains(string(body), `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+				t.Fatalf("credits retry missing enabledCreditTypes: %s", string(body))
+			}
+			w.WriteHeader(http.StatusForbidden)
+			_, _ = w.Write([]byte(`{"error":{"message":"permission denied"}}`))
+		default:
+			t.Fatalf("unexpected first server request count %d", reqNum)
+		}
+	}))
+	defer firstServer.Close()
+
+	secondServer := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		mu.Lock()
+		secondCount++
+		mu.Unlock()
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"response":{"candidates":[{"content":{"role":"model","parts":[{"text":"ok"}]}}],"usageMetadata":{"promptTokenCount":1,"candidatesTokenCount":1,"totalTokenCount":2}}}`))
+	}))
+	defer secondServer.Close()
+
+	exec := NewAntigravityExecutor(&config.Config{
+		QuotaExceeded: config.QuotaExceeded{AntigravityCredits: true},
+	})
+	auth := &cliproxyauth.Auth{
+		ID: "auth-baseurl-fallback",
+		Attributes: map[string]string{
+			"base_url": firstServer.URL,
+		},
+		Metadata: map[string]any{
+			"access_token": "token",
+			"project_id":   "project-1",
+			"expired":      time.Now().Add(1 * time.Hour).Format(time.RFC3339),
+		},
+	}
+
+	originalOrder := antigravityBaseURLFallbackOrder
+	defer func() { antigravityBaseURLFallbackOrder = originalOrder }()
+	antigravityBaseURLFallbackOrder = func(auth *cliproxyauth.Auth) []string {
+		return []string{firstServer.URL, secondServer.URL}
+	}
+
+	resp, err := exec.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gemini-2.5-flash",
+		Payload: []byte(`{"request":{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FormatAntigravity,
+	})
+	if err != nil {
+		t.Fatalf("Execute() error = %v", err)
+	}
+	if len(resp.Payload) == 0 {
+		t.Fatal("Execute() returned empty payload")
+	}
+	if firstCount != 2 {
+		t.Fatalf("first server request count = %d, want 2", firstCount)
+	}
+	if secondCount != 1 {
+		t.Fatalf("second server request count = %d, want 1", secondCount)
+	}
+}
+
+func TestAntigravityExecute_DoesNotDirectInjectCreditsWhenFlagDisabled(t *testing.T) {
+	resetAntigravityCreditsRetryState()
+	t.Cleanup(resetAntigravityCreditsRetryState)
+
+	var requestBodies []string
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		_ = r.Body.Close()
+		requestBodies = append(requestBodies, string(body))
+		w.WriteHeader(http.StatusTooManyRequests)
+		_, _ = w.Write([]byte(`{"error":{"status":"RESOURCE_EXHAUSTED","message":"QUOTA_EXHAUSTED"}}`))
+	}))
+	defer server.Close()
+
+	exec := NewAntigravityExecutor(&config.Config{
+		QuotaExceeded: config.QuotaExceeded{AntigravityCredits: false},
+	})
+	auth := &cliproxyauth.Auth{
+		ID: "auth-flag-disabled",
+		Attributes: map[string]string{
+			"base_url": server.URL,
+		},
+		Metadata: map[string]any{
+			"access_token": "token",
+			"project_id":   "project-1",
+			"expired":      time.Now().Add(1 * time.Hour).Format(time.RFC3339),
+		},
+	}
+	markAntigravityPreferCredits(auth, "gemini-2.5-flash", time.Now(), nil)
+
+	_, err := exec.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gemini-2.5-flash",
+		Payload: []byte(`{"request":{"contents":[{"role":"user","parts":[{"text":"hi"}]}]}}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FormatAntigravity,
+	})
+	if err == nil {
+		t.Fatal("Execute() error = nil, want 429")
+	}
+	if len(requestBodies) != 1 {
+		t.Fatalf("request count = %d, want 1", len(requestBodies))
+	}
+	if strings.Contains(requestBodies[0], `"enabledCreditTypes":["GOOGLE_ONE_AI"]`) {
+		t.Fatalf("request unexpectedly used enabledCreditTypes with flag disabled: %s", requestBodies[0])
+	}
+}

internal/runtime/executor/claude_executor.go

@@ -6,7 +6,6 @@ import (
 	"compress/flate"
 	"compress/gzip"
 	"context"
-	"crypto/rand"
 	"crypto/sha256"
 	"encoding/hex"
 	"encoding/json"
@@ -18,10 +17,13 @@ import (
 	"time"
 
 	"github.com/andybalholm/brotli"
+	"github.com/google/uuid"
 	"github.com/klauspost/compress/zstd"
 	claudeauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/claude"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/misc"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -44,6 +46,10 @@ type ClaudeExecutor struct {
 // Previously "proxy_" was used but this is a detectable fingerprint difference.
 const claudeToolPrefix = ""
 
+// Anthropic-compatible upstreams may reject or even crash when Claude models
+// omit max_tokens. Prefer registered model metadata before using a fallback.
+const defaultModelMaxTokens = 1024
+
 func NewClaudeExecutor(cfg *config.Config) *ClaudeExecutor { return &ClaudeExecutor{cfg: cfg} }
 
 func (e *ClaudeExecutor) Identifier() string { return "claude" }
@@ -86,7 +92,7 @@ func (e *ClaudeExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Aut
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewUtlsHTTPClient(e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -101,8 +107,8 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 		baseURL = "https://api.anthropic.com"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("claude")
 	// Use streaming translation to preserve function calling, except for claude.
@@ -125,11 +131,13 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 	// based on client type and configuration.
 	body = applyCloaking(ctx, e.cfg, auth, body, baseModel, apiKey)
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	body = ensureModelMaxTokens(body, baseModel)
 
 	// Disable thinking if tool_choice forces tool use (Anthropic API constraint)
 	body = disableThinkingIfToolChoiceForced(body)
+	body = normalizeClaudeTemperatureForThinking(body)
 
 	// Auto-inject cache_control if missing (optimization for ClawdBot/clients without caching support)
 	if countCacheControls(body) == 0 {
@@ -153,6 +161,9 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 	if isClaudeOAuthToken(apiKey) && !auth.ToolPrefixDisabled() {
 		bodyForUpstream = applyClaudeToolPrefix(body, claudeToolPrefix)
 	}
+	if experimentalCCHSigningEnabled(e.cfg, auth) {
+		bodyForUpstream = signAnthropicMessagesBody(bodyForUpstream)
+	}
 
 	url := fmt.Sprintf("%s/v1/messages?beta=true", baseURL)
 	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(bodyForUpstream))
@@ -166,7 +177,7 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -178,33 +189,33 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewUtlsHTTPClient(e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		// Decompress error responses — pass the Content-Encoding value (may be empty)
 		// and let decodeResponseBody handle both header-declared and magic-byte-detected
 		// compression.  This keeps error-path behaviour consistent with the success path.
 		errBody, decErr := decodeResponseBody(httpResp.Body, httpResp.Header.Get("Content-Encoding"))
 		if decErr != nil {
-			recordAPIResponseError(ctx, e.cfg, decErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, decErr)
 			msg := fmt.Sprintf("failed to decode error response body: %v", decErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			return resp, statusErr{code: httpResp.StatusCode, msg: msg}
 		}
 		b, readErr := io.ReadAll(errBody)
 		if readErr != nil {
-			recordAPIResponseError(ctx, e.cfg, readErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, readErr)
 			msg := fmt.Sprintf("failed to read error response body: %v", readErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			b = []byte(msg)
 		}
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		if errClose := errBody.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
@@ -213,7 +224,7 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 	}
 	decodedBody, err := decodeResponseBody(httpResp.Body, httpResp.Header.Get("Content-Encoding"))
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
 		}
@@ -226,19 +237,19 @@ func (e *ClaudeExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 	}()
 	data, err := io.ReadAll(decodedBody)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 	if stream {
 		lines := bytes.Split(data, []byte("\n"))
 		for _, line := range lines {
-			if detail, ok := parseClaudeStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			if detail, ok := helps.ParseClaudeStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 		}
 	} else {
-		reporter.publish(ctx, parseClaudeUsage(data))
+		reporter.Publish(ctx, helps.ParseClaudeUsage(data))
 	}
 	if isClaudeOAuthToken(apiKey) && !auth.ToolPrefixDisabled() {
 		data = stripClaudeToolPrefixFromResponse(data, claudeToolPrefix)
@@ -269,8 +280,8 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		baseURL = "https://api.anthropic.com"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("claude")
 	originalPayloadSource := req.Payload
@@ -291,11 +302,13 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 	// based on client type and configuration.
 	body = applyCloaking(ctx, e.cfg, auth, body, baseModel, apiKey)
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	body = ensureModelMaxTokens(body, baseModel)
 
 	// Disable thinking if tool_choice forces tool use (Anthropic API constraint)
 	body = disableThinkingIfToolChoiceForced(body)
+	body = normalizeClaudeTemperatureForThinking(body)
 
 	// Auto-inject cache_control if missing (optimization for ClawdBot/clients without caching support)
 	if countCacheControls(body) == 0 {
@@ -316,6 +329,9 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 	if isClaudeOAuthToken(apiKey) && !auth.ToolPrefixDisabled() {
 		bodyForUpstream = applyClaudeToolPrefix(body, claudeToolPrefix)
 	}
+	if experimentalCCHSigningEnabled(e.cfg, auth) {
+		bodyForUpstream = signAnthropicMessagesBody(bodyForUpstream)
+	}
 
 	url := fmt.Sprintf("%s/v1/messages?beta=true", baseURL)
 	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(bodyForUpstream))
@@ -329,7 +345,7 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -341,33 +357,33 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewUtlsHTTPClient(e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		// Decompress error responses — pass the Content-Encoding value (may be empty)
 		// and let decodeResponseBody handle both header-declared and magic-byte-detected
 		// compression.  This keeps error-path behaviour consistent with the success path.
 		errBody, decErr := decodeResponseBody(httpResp.Body, httpResp.Header.Get("Content-Encoding"))
 		if decErr != nil {
-			recordAPIResponseError(ctx, e.cfg, decErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, decErr)
 			msg := fmt.Sprintf("failed to decode error response body: %v", decErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			return nil, statusErr{code: httpResp.StatusCode, msg: msg}
 		}
 		b, readErr := io.ReadAll(errBody)
 		if readErr != nil {
-			recordAPIResponseError(ctx, e.cfg, readErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, readErr)
 			msg := fmt.Sprintf("failed to read error response body: %v", readErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			b = []byte(msg)
 		}
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := errBody.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
 		}
@@ -376,7 +392,7 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 	}
 	decodedBody, err := decodeResponseBody(httpResp.Body, httpResp.Header.Get("Content-Encoding"))
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
 		}
@@ -397,9 +413,9 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 			scanner.Buffer(nil, 52_428_800) // 50MB
 			for scanner.Scan() {
 				line := scanner.Bytes()
-				appendAPIResponseChunk(ctx, e.cfg, line)
-				if detail, ok := parseClaudeStreamUsage(line); ok {
-					reporter.publish(ctx, detail)
+				helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+				if detail, ok := helps.ParseClaudeStreamUsage(line); ok {
+					reporter.Publish(ctx, detail)
 				}
 				if isClaudeOAuthToken(apiKey) && !auth.ToolPrefixDisabled() {
 					line = stripClaudeToolPrefixFromStreamLine(line, claudeToolPrefix)
@@ -411,8 +427,8 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 				out <- cliproxyexecutor.StreamChunk{Payload: cloned}
 			}
 			if errScan := scanner.Err(); errScan != nil {
-				recordAPIResponseError(ctx, e.cfg, errScan)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+				reporter.PublishFailure(ctx)
 				out <- cliproxyexecutor.StreamChunk{Err: errScan}
 			}
 			return
@@ -424,9 +440,9 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseClaudeStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseClaudeStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			if isClaudeOAuthToken(apiKey) && !auth.ToolPrefixDisabled() {
 				line = stripClaudeToolPrefixFromStreamLine(line, claudeToolPrefix)
@@ -446,8 +462,8 @@ func (e *ClaudeExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 			}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -496,7 +512,7 @@ func (e *ClaudeExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -508,32 +524,32 @@ func (e *ClaudeExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewUtlsHTTPClient(e.cfg, auth, 0)
 	resp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return cliproxyexecutor.Response{}, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
 	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
 		// Decompress error responses — pass the Content-Encoding value (may be empty)
 		// and let decodeResponseBody handle both header-declared and magic-byte-detected
 		// compression.  This keeps error-path behaviour consistent with the success path.
 		errBody, decErr := decodeResponseBody(resp.Body, resp.Header.Get("Content-Encoding"))
 		if decErr != nil {
-			recordAPIResponseError(ctx, e.cfg, decErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, decErr)
 			msg := fmt.Sprintf("failed to decode error response body: %v", decErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			return cliproxyexecutor.Response{}, statusErr{code: resp.StatusCode, msg: msg}
 		}
 		b, readErr := io.ReadAll(errBody)
 		if readErr != nil {
-			recordAPIResponseError(ctx, e.cfg, readErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, readErr)
 			msg := fmt.Sprintf("failed to read error response body: %v", readErr)
-			logWithRequestID(ctx).Warn(msg)
+			helps.LogWithRequestID(ctx).Warn(msg)
 			b = []byte(msg)
 		}
-		appendAPIResponseChunk(ctx, e.cfg, b)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
 		if errClose := errBody.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
 		}
@@ -541,7 +557,7 @@ func (e *ClaudeExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 	}
 	decodedBody, err := decodeResponseBody(resp.Body, resp.Header.Get("Content-Encoding"))
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		if errClose := resp.Body.Close(); errClose != nil {
 			log.Errorf("response body close error: %v", errClose)
 		}
@@ -554,10 +570,10 @@ func (e *ClaudeExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 	}()
 	data, err := io.ReadAll(decodedBody)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return cliproxyexecutor.Response{}, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 	count := gjson.GetBytes(data, "input_tokens").Int()
 	out := sdktranslator.TranslateTokenCount(ctx, to, from, count, data)
 	return cliproxyexecutor.Response{Payload: out, Headers: resp.Header.Clone()}, nil
@@ -637,6 +653,25 @@ func disableThinkingIfToolChoiceForced(body []byte) []byte {
 	return body
 }
 
+// normalizeClaudeTemperatureForThinking keeps Anthropic message requests valid when
+// thinking is enabled. Anthropic rejects temperatures other than 1 when
+// thinking.type is enabled/adaptive/auto.
+func normalizeClaudeTemperatureForThinking(body []byte) []byte {
+	if !gjson.GetBytes(body, "temperature").Exists() {
+		return body
+	}
+
+	thinkingType := strings.ToLower(strings.TrimSpace(gjson.GetBytes(body, "thinking.type").String()))
+	switch thinkingType {
+	case "enabled", "adaptive", "auto":
+		if temp := gjson.GetBytes(body, "temperature"); temp.Exists() && temp.Type == gjson.Number && temp.Float() == 1 {
+			return body
+		}
+		body, _ = sjson.SetBytes(body, "temperature", 1)
+	}
+	return body
+}
+
 type compositeReadCloser struct {
 	io.Reader
 	closers []func() error
@@ -793,13 +828,13 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string,
 	if ginCtx, ok := r.Context().Value("gin").(*gin.Context); ok && ginCtx != nil && ginCtx.Request != nil {
 		ginHeaders = ginCtx.Request.Header
 	}
-	stabilizeDeviceProfile := claudeDeviceProfileStabilizationEnabled(cfg)
-	var deviceProfile claudeDeviceProfile
+	stabilizeDeviceProfile := helps.ClaudeDeviceProfileStabilizationEnabled(cfg)
+	var deviceProfile helps.ClaudeDeviceProfile
 	if stabilizeDeviceProfile {
-		deviceProfile = resolveClaudeDeviceProfile(auth, apiKey, ginHeaders, cfg)
+		deviceProfile = helps.ResolveClaudeDeviceProfile(auth, apiKey, ginHeaders, cfg)
 	}
 
-	baseBetas := "claude-code-20250219,oauth-2025-04-20,interleaved-thinking-2025-05-14,context-management-2025-06-27,prompt-caching-scope-2026-01-05"
+	baseBetas := "claude-code-20250219,oauth-2025-04-20,interleaved-thinking-2025-05-14,context-management-2025-06-27,prompt-caching-scope-2026-01-05,structured-outputs-2025-12-15,fast-mode-2026-02-01,redact-thinking-2026-02-12,token-efficient-tools-2026-03-28"
 	if val := strings.TrimSpace(ginHeaders.Get("Anthropic-Beta")); val != "" {
 		baseBetas = val
 		if !strings.Contains(val, "oauth") {
@@ -813,6 +848,14 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string,
 			hasClaude1MHeader = true
 		}
 	}
+	// Also check auth attributes — GitLab Duo sets gitlab_duo_force_context_1m
+	// when routing through the Anthropic gateway, but the gin headers won't have
+	// X-CPA-CLAUDE-1M because the request is internally constructed.
+	if !hasClaude1MHeader && auth != nil && auth.Attributes != nil {
+		if auth.Attributes["gitlab_duo_force_context_1m"] == "true" {
+			hasClaude1MHeader = true
+		}
+	}
 
 	// Merge extra betas from request body and request flags.
 	if len(extraBetas) > 0 || hasClaude1MHeader {
@@ -837,13 +880,22 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string,
 	r.Header.Set("Anthropic-Beta", baseBetas)
 
 	misc.EnsureHeader(r.Header, ginHeaders, "Anthropic-Version", "2023-06-01")
-	misc.EnsureHeader(r.Header, ginHeaders, "Anthropic-Dangerous-Direct-Browser-Access", "true")
+	// Only set browser access header for API key mode; real Claude Code CLI does not send it.
+	if useAPIKey {
+		misc.EnsureHeader(r.Header, ginHeaders, "Anthropic-Dangerous-Direct-Browser-Access", "true")
+	}
 	misc.EnsureHeader(r.Header, ginHeaders, "X-App", "cli")
 	// Values below match Claude Code 2.1.63 / @anthropic-ai/sdk 0.74.0 (updated 2026-02-28).
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Stainless-Retry-Count", "0")
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Stainless-Runtime", "node")
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Stainless-Lang", "js")
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Stainless-Timeout", hdrDefault(hd.Timeout, "600"))
+	// Session ID: stable per auth/apiKey, matches Claude Code's X-Claude-Code-Session-Id header.
+	misc.EnsureHeader(r.Header, ginHeaders, "X-Claude-Code-Session-Id", helps.CachedSessionID(apiKey))
+	// Per-request UUID, matches Claude Code's x-client-request-id for first-party API.
+	if isAnthropicBase {
+		misc.EnsureHeader(r.Header, ginHeaders, "x-client-request-id", uuid.New().String())
+	}
 	r.Header.Set("Connection", "keep-alive")
 	if stream {
 		r.Header.Set("Accept", "text/event-stream")
@@ -858,16 +910,16 @@ func applyClaudeHeaders(r *http.Request, auth *cliproxyauth.Auth, apiKey string,
 	// Legacy mode keeps OS/Arch runtime-derived; stabilized mode pins OS/Arch
 	// to the configured baseline while still allowing newer official
 	// User-Agent/package/runtime tuples to upgrade the software fingerprint.
+	if stabilizeDeviceProfile {
+		helps.ApplyClaudeDeviceProfileHeaders(r, deviceProfile)
+	} else {
+		helps.ApplyClaudeLegacyDeviceHeaders(r, ginHeaders, cfg)
+	}
 	var attrs map[string]string
 	if auth != nil {
 		attrs = auth.Attributes
 	}
 	util.ApplyCustomHeadersFromAttrs(r, attrs)
-	if stabilizeDeviceProfile {
-		applyClaudeDeviceProfileHeaders(r, deviceProfile)
-	} else {
-		applyClaudeLegacyDeviceHeaders(r, ginHeaders, cfg)
-	}
 	// Re-enforce Accept-Encoding: identity after ApplyCustomHeadersFromAttrs, which
 	// may override it with a user-configured value.  Compressed SSE breaks the line
 	// scanner regardless of user preference, so this is non-negotiable for streams.
@@ -893,7 +945,7 @@ func claudeCreds(a *cliproxyauth.Auth) (apiKey, baseURL string) {
 }
 
 func checkSystemInstructions(payload []byte) []byte {
-	return checkSystemInstructionsWithMode(payload, false)
+	return checkSystemInstructionsWithSigningMode(payload, false, false, "2.1.63", "", "")
 }
 
 func isClaudeOAuthToken(apiKey string) bool {
@@ -1037,7 +1089,7 @@ func stripClaudeToolPrefixFromStreamLine(line []byte, prefix string) []byte {
 	if prefix == "" {
 		return line
 	}
-	payload := jsonPayload(line)
+	payload := helps.JSONPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return line
 	}
@@ -1088,6 +1140,38 @@ func getClientUserAgent(ctx context.Context) string {
 	return ""
 }
 
+// parseEntrypointFromUA extracts the entrypoint from a Claude Code User-Agent.
+// Format: "claude-cli/x.y.z (external, cli)" → "cli"
+// Format: "claude-cli/x.y.z (external, vscode)" → "vscode"
+// Returns "cli" if parsing fails or UA is not Claude Code.
+func parseEntrypointFromUA(userAgent string) string {
+	// Find content inside parentheses
+	start := strings.Index(userAgent, "(")
+	end := strings.LastIndex(userAgent, ")")
+	if start < 0 || end <= start {
+		return "cli"
+	}
+	inner := userAgent[start+1 : end]
+	// Split by comma, take the second part (entrypoint is at index 1, after USER_TYPE)
+	// Format: "(USER_TYPE, ENTRYPOINT[, extra...])"
+	parts := strings.Split(inner, ",")
+	if len(parts) >= 2 {
+		ep := strings.TrimSpace(parts[1])
+		if ep != "" {
+			return ep
+		}
+	}
+	return "cli"
+}
+
+// getWorkloadFromContext extracts workload identifier from the gin request headers.
+func getWorkloadFromContext(ctx context.Context) string {
+	if ginCtx, ok := ctx.Value("gin").(*gin.Context); ok && ginCtx != nil && ginCtx.Request != nil {
+		return strings.TrimSpace(ginCtx.GetHeader("X-CPA-Claude-Workload"))
+	}
+	return ""
+}
+
 // getCloakConfigFromAuth extracts cloak configuration from auth attributes.
 // Returns (cloakMode, strictMode, sensitiveWords, cacheUserID).
 func getCloakConfigFromAuth(auth *cliproxyauth.Auth) (string, bool, []string, bool) {
@@ -1115,43 +1199,14 @@ func getCloakConfigFromAuth(auth *cliproxyauth.Auth) (string, bool, []string, bo
 	return cloakMode, strictMode, sensitiveWords, cacheUserID
 }
 
-// resolveClaudeKeyCloakConfig finds the matching ClaudeKey config and returns its CloakConfig.
-func resolveClaudeKeyCloakConfig(cfg *config.Config, auth *cliproxyauth.Auth) *config.CloakConfig {
-	if cfg == nil || auth == nil {
-		return nil
-	}
-
-	apiKey, baseURL := claudeCreds(auth)
-	if apiKey == "" {
-		return nil
-	}
-
-	for i := range cfg.ClaudeKey {
-		entry := &cfg.ClaudeKey[i]
-		cfgKey := strings.TrimSpace(entry.APIKey)
-		cfgBase := strings.TrimSpace(entry.BaseURL)
-
-		// Match by API key
-		if strings.EqualFold(cfgKey, apiKey) {
-			// If baseURL is specified, also check it
-			if baseURL != "" && cfgBase != "" && !strings.EqualFold(cfgBase, baseURL) {
-				continue
-			}
-			return entry.Cloak
-		}
-	}
-
-	return nil
-}
-
 // injectFakeUserID generates and injects a fake user ID into the request metadata.
 // When useCache is false, a new user ID is generated for every call.
 func injectFakeUserID(payload []byte, apiKey string, useCache bool) []byte {
 	generateID := func() string {
 		if useCache {
-			return cachedUserID(apiKey)
+			return helps.CachedUserID(apiKey)
 		}
-		return generateFakeUserID()
+		return helps.GenerateFakeUserID()
 	}
 
 	metadata := gjson.GetBytes(payload, "metadata")
@@ -1161,38 +1216,84 @@ func injectFakeUserID(payload []byte, apiKey string, useCache bool) []byte {
 	}
 
 	existingUserID := gjson.GetBytes(payload, "metadata.user_id").String()
-	if existingUserID == "" || !isValidUserID(existingUserID) {
+	if existingUserID == "" || !helps.IsValidUserID(existingUserID) {
 		payload, _ = sjson.SetBytes(payload, "metadata.user_id", generateID())
 	}
 	return payload
 }
 
-// generateBillingHeader creates the x-anthropic-billing-header text block that
-// real Claude Code prepends to every system prompt array.
-// Format: x-anthropic-billing-header: cc_version=<ver>.<build>; cc_entrypoint=cli; cch=<hash>;
-func generateBillingHeader(payload []byte) string {
-	// Generate a deterministic cch hash from the payload content (system + messages + tools).
-	// Real Claude Code uses a 5-char hex hash that varies per request.
-	h := sha256.Sum256(payload)
-	cch := hex.EncodeToString(h[:])[:5]
+// fingerprintSalt is the salt used by Claude Code to compute the 3-char build fingerprint.
+const fingerprintSalt = "59cf53e54c78"
 
-	// Build hash: 3-char hex, matches the pattern seen in real requests (e.g. "a43")
-	buildBytes := make([]byte, 2)
-	_, _ = rand.Read(buildBytes)
-	buildHash := hex.EncodeToString(buildBytes)[:3]
-
-	return fmt.Sprintf("x-anthropic-billing-header: cc_version=2.1.63.%s; cc_entrypoint=cli; cch=%s;", buildHash, cch)
+// computeFingerprint computes the 3-char build fingerprint that Claude Code embeds in cc_version.
+// Algorithm: SHA256(salt + messageText[4] + messageText[7] + messageText[20] + version)[:3]
+func computeFingerprint(messageText, version string) string {
+	indices := [3]int{4, 7, 20}
+	runes := []rune(messageText)
+	var sb strings.Builder
+	for _, idx := range indices {
+		if idx < len(runes) {
+			sb.WriteRune(runes[idx])
+		} else {
+			sb.WriteRune('0')
+		}
+	}
+	input := fingerprintSalt + sb.String() + version
+	h := sha256.Sum256([]byte(input))
+	return hex.EncodeToString(h[:])[:3]
 }
 
-// checkSystemInstructionsWithMode injects Claude Code-style system blocks:
+// generateBillingHeader creates the x-anthropic-billing-header text block that
+// real Claude Code prepends to every system prompt array.
+// Format: x-anthropic-billing-header: cc_version=<ver>.<build>; cc_entrypoint=<ep>; cch=<hash>; [cc_workload=<wl>;]
+func generateBillingHeader(payload []byte, experimentalCCHSigning bool, version, messageText, entrypoint, workload string) string {
+	if entrypoint == "" {
+		entrypoint = "cli"
+	}
+	buildHash := computeFingerprint(messageText, version)
+	workloadPart := ""
+	if workload != "" {
+		workloadPart = fmt.Sprintf(" cc_workload=%s;", workload)
+	}
+
+	if experimentalCCHSigning {
+		return fmt.Sprintf("x-anthropic-billing-header: cc_version=%s.%s; cc_entrypoint=%s; cch=00000;%s", version, buildHash, entrypoint, workloadPart)
+	}
+
+	// Generate a deterministic cch hash from the payload content (system + messages + tools).
+	h := sha256.Sum256(payload)
+	cch := hex.EncodeToString(h[:])[:5]
+	return fmt.Sprintf("x-anthropic-billing-header: cc_version=%s.%s; cc_entrypoint=%s; cch=%s;%s", version, buildHash, entrypoint, cch, workloadPart)
+}
+
+func checkSystemInstructionsWithMode(payload []byte, strictMode bool) []byte {
+	return checkSystemInstructionsWithSigningMode(payload, strictMode, false, "2.1.63", "", "")
+}
+
+// checkSystemInstructionsWithSigningMode injects Claude Code-style system blocks:
 //
 //	system[0]: billing header (no cache_control)
 //	system[1]: agent identifier (no cache_control)
 //	system[2..]: user system messages (cache_control added when missing)
-func checkSystemInstructionsWithMode(payload []byte, strictMode bool) []byte {
+func checkSystemInstructionsWithSigningMode(payload []byte, strictMode bool, experimentalCCHSigning bool, version, entrypoint, workload string) []byte {
 	system := gjson.GetBytes(payload, "system")
 
-	billingText := generateBillingHeader(payload)
+	// Extract original message text for fingerprint computation (before billing injection).
+	// Use the first system text block's content as the fingerprint source.
+	messageText := ""
+	if system.IsArray() {
+		system.ForEach(func(_, part gjson.Result) bool {
+			if part.Get("type").String() == "text" {
+				messageText = part.Get("text").String()
+				return false
+			}
+			return true
+		})
+	} else if system.Type == gjson.String {
+		messageText = system.String()
+	}
+
+	billingText := generateBillingHeader(payload, experimentalCCHSigning, version, messageText, entrypoint, workload)
 	billingBlock := fmt.Sprintf(`{"type":"text","text":"%s"}`, billingText)
 	// No cache_control on the agent block. It is a cloaking artifact with zero cache
 	// value (the last system block is what actually triggers caching of all system content).
@@ -1247,51 +1348,44 @@ func checkSystemInstructionsWithMode(payload []byte, strictMode bool) []byte {
 // Cloaking includes: system prompt injection, fake user ID, and sensitive word obfuscation.
 func applyCloaking(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, payload []byte, model string, apiKey string) []byte {
 	clientUserAgent := getClientUserAgent(ctx)
+	useExperimentalCCHSigning := experimentalCCHSigningEnabled(cfg, auth)
 
 	// Get cloak config from ClaudeKey configuration
 	cloakCfg := resolveClaudeKeyCloakConfig(cfg, auth)
+	attrMode, attrStrict, attrWords, attrCache := getCloakConfigFromAuth(auth)
 
 	// Determine cloak settings
-	var cloakMode string
-	var strictMode bool
-	var sensitiveWords []string
-	var cacheUserID bool
+	cloakMode := attrMode
+	strictMode := attrStrict
+	sensitiveWords := attrWords
+	cacheUserID := attrCache
 
 	if cloakCfg != nil {
-		cloakMode = cloakCfg.Mode
-		strictMode = cloakCfg.StrictMode
-		sensitiveWords = cloakCfg.SensitiveWords
+		if mode := strings.TrimSpace(cloakCfg.Mode); mode != "" {
+			cloakMode = mode
+		}
+		if cloakCfg.StrictMode {
+			strictMode = true
+		}
+		if len(cloakCfg.SensitiveWords) > 0 {
+			sensitiveWords = cloakCfg.SensitiveWords
+		}
 		if cloakCfg.CacheUserID != nil {
 			cacheUserID = *cloakCfg.CacheUserID
 		}
 	}
 
-	// Fallback to auth attributes if no config found
-	if cloakMode == "" {
-		attrMode, attrStrict, attrWords, attrCache := getCloakConfigFromAuth(auth)
-		cloakMode = attrMode
-		if !strictMode {
-			strictMode = attrStrict
-		}
-		if len(sensitiveWords) == 0 {
-			sensitiveWords = attrWords
-		}
-		if cloakCfg == nil || cloakCfg.CacheUserID == nil {
-			cacheUserID = attrCache
-		}
-	} else if cloakCfg == nil || cloakCfg.CacheUserID == nil {
-		_, _, _, attrCache := getCloakConfigFromAuth(auth)
-		cacheUserID = attrCache
-	}
-
 	// Determine if cloaking should be applied
-	if !shouldCloak(cloakMode, clientUserAgent) {
+	if !helps.ShouldCloak(cloakMode, clientUserAgent) {
 		return payload
 	}
 
 	// Skip system instructions for claude-3-5-haiku models
 	if !strings.HasPrefix(model, "claude-3-5-haiku") {
-		payload = checkSystemInstructionsWithMode(payload, strictMode)
+		billingVersion := helps.DefaultClaudeVersion(cfg)
+		entrypoint := parseEntrypointFromUA(clientUserAgent)
+		workload := getWorkloadFromContext(ctx)
+		payload = checkSystemInstructionsWithSigningMode(payload, strictMode, useExperimentalCCHSigning, billingVersion, entrypoint, workload)
 	}
 
 	// Inject fake user ID
@@ -1299,8 +1393,8 @@ func applyCloaking(ctx context.Context, cfg *config.Config, auth *cliproxyauth.A
 
 	// Apply sensitive word obfuscation
 	if len(sensitiveWords) > 0 {
-		matcher := buildSensitiveWordMatcher(sensitiveWords)
-		payload = obfuscateSensitiveWords(payload, matcher)
+		matcher := helps.BuildSensitiveWordMatcher(sensitiveWords)
+		payload = helps.ObfuscateSensitiveWords(payload, matcher)
 	}
 
 	return payload
@@ -1310,7 +1404,7 @@ func applyCloaking(ctx context.Context, cfg *config.Config, auth *cliproxyauth.A
 // According to Anthropic's documentation, cache prefixes are created in order: tools -> system -> messages.
 // This function adds cache_control to:
 // 1. The LAST tool in the tools array (caches all tool definitions)
-// 2. The LAST element in the system array (caches system prompt)
+// 2. The LAST system prompt element
 // 3. The SECOND-TO-LAST user turn (caches conversation history for multi-turn)
 //
 // Up to 4 cache breakpoints are allowed per request. Tools, System, and Messages are INDEPENDENT breakpoints.
@@ -1880,3 +1974,26 @@ func injectSystemCacheControl(payload []byte) []byte {
 
 	return payload
 }
+
+func ensureModelMaxTokens(body []byte, modelID string) []byte {
+	if len(body) == 0 || !gjson.ValidBytes(body) {
+		return body
+	}
+
+	if maxTokens := gjson.GetBytes(body, "max_tokens"); maxTokens.Exists() {
+		return body
+	}
+
+	for _, provider := range registry.GetGlobalRegistry().GetModelProviders(strings.TrimSpace(modelID)) {
+		if strings.EqualFold(provider, "claude") {
+			maxTokens := defaultModelMaxTokens
+			if info := registry.GetGlobalRegistry().GetModelInfo(strings.TrimSpace(modelID), "claude"); info != nil && info.MaxCompletionTokens > 0 {
+				maxTokens = info.MaxCompletionTokens
+			}
+			body, _ = sjson.SetBytes(body, "max_tokens", maxTokens)
+			return body
+		}
+	}
+
+	return body
+}

internal/runtime/executor/claude_executor_test.go

@@ -4,9 +4,11 @@ import (
 	"bytes"
 	"compress/gzip"
 	"context"
+	"fmt"
 	"io"
 	"net/http"
 	"net/http/httptest"
+	"regexp"
 	"strings"
 	"sync"
 	"testing"
@@ -14,7 +16,10 @@ import (
 
 	"github.com/gin-gonic/gin"
 	"github.com/klauspost/compress/zstd"
+	xxHash64 "github.com/pierrec/xxHash/xxHash64"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
@@ -23,9 +28,7 @@ import (
 )
 
 func resetClaudeDeviceProfileCache() {
-	claudeDeviceProfileCacheMu.Lock()
-	claudeDeviceProfileCache = make(map[string]claudeDeviceProfileCacheEntry)
-	claudeDeviceProfileCacheMu.Unlock()
+	helps.ResetClaudeDeviceProfileCache()
 }
 
 func newClaudeHeaderTestRequest(t *testing.T, incoming http.Header) *http.Request {
@@ -98,7 +101,7 @@ func TestApplyClaudeHeaders_UsesConfiguredBaselineFingerprint(t *testing.T) {
 	req := newClaudeHeaderTestRequest(t, incoming)
 	applyClaudeHeaders(req, auth, "key-baseline", false, nil, cfg)
 
-	assertClaudeFingerprint(t, req.Header, "claude-cli/2.1.70 (external, cli)", "0.80.0", "v24.5.0", "MacOS", "arm64")
+	assertClaudeFingerprint(t, req.Header, "evil-client/9.9", "9.9.9", "v24.5.0", "Linux", "x64")
 	if got := req.Header.Get("X-Stainless-Timeout"); got != "900" {
 		t.Fatalf("X-Stainless-Timeout = %q, want %q", got, "900")
 	}
@@ -338,7 +341,7 @@ func TestResolveClaudeDeviceProfile_RechecksCacheBeforeStoringCandidate(t *testi
 	var pauseOnce sync.Once
 	var releaseOnce sync.Once
 
-	claudeDeviceProfileBeforeCandidateStore = func(candidate claudeDeviceProfile) {
+	helps.ClaudeDeviceProfileBeforeCandidateStore = func(candidate helps.ClaudeDeviceProfile) {
 		if candidate.UserAgent != "claude-cli/2.1.62 (external, cli)" {
 			return
 		}
@@ -346,13 +349,13 @@ func TestResolveClaudeDeviceProfile_RechecksCacheBeforeStoringCandidate(t *testi
 		<-releaseLow
 	}
 	t.Cleanup(func() {
-		claudeDeviceProfileBeforeCandidateStore = nil
+		helps.ClaudeDeviceProfileBeforeCandidateStore = nil
 		releaseOnce.Do(func() { close(releaseLow) })
 	})
 
-	lowResultCh := make(chan claudeDeviceProfile, 1)
+	lowResultCh := make(chan helps.ClaudeDeviceProfile, 1)
 	go func() {
-		lowResultCh <- resolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
+		lowResultCh <- helps.ResolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
 			"User-Agent":                  []string{"claude-cli/2.1.62 (external, cli)"},
 			"X-Stainless-Package-Version": []string{"0.74.0"},
 			"X-Stainless-Runtime-Version": []string{"v24.3.0"},
@@ -367,7 +370,7 @@ func TestResolveClaudeDeviceProfile_RechecksCacheBeforeStoringCandidate(t *testi
 		t.Fatal("timed out waiting for lower candidate to pause before storing")
 	}
 
-	highResult := resolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
+	highResult := helps.ResolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
 		"User-Agent":                  []string{"claude-cli/2.1.63 (external, cli)"},
 		"X-Stainless-Package-Version": []string{"0.75.0"},
 		"X-Stainless-Runtime-Version": []string{"v24.4.0"},
@@ -398,7 +401,7 @@ func TestResolveClaudeDeviceProfile_RechecksCacheBeforeStoringCandidate(t *testi
 		t.Fatalf("highResult platform = %s/%s, want %s/%s", highResult.OS, highResult.Arch, "MacOS", "arm64")
 	}
 
-	cached := resolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
+	cached := helps.ResolveClaudeDeviceProfile(auth, "key-racy-upgrade", http.Header{
 		"User-Agent": []string{"curl/8.7.1"},
 	}, cfg)
 	if cached.UserAgent != "claude-cli/2.1.63 (external, cli)" {
@@ -564,7 +567,7 @@ func TestApplyClaudeHeaders_LegacyModeFallsBackToRuntimeOSArchWhenMissing(t *tes
 	})
 	applyClaudeHeaders(req, auth, "key-legacy-runtime-os-arch", false, nil, cfg)
 
-	assertClaudeFingerprint(t, req.Header, "claude-cli/2.1.60 (external, cli)", "0.70.0", "v22.0.0", mapStainlessOS(), mapStainlessArch())
+	assertClaudeFingerprint(t, req.Header, "claude-cli/2.1.60 (external, cli)", "0.70.0", "v22.0.0", helps.MapStainlessOS(), helps.MapStainlessArch())
 }
 
 func TestApplyClaudeHeaders_UnsetStabilizationAlsoUsesLegacyRuntimeOSArchFallback(t *testing.T) {
@@ -591,14 +594,14 @@ func TestApplyClaudeHeaders_UnsetStabilizationAlsoUsesLegacyRuntimeOSArchFallbac
 	})
 	applyClaudeHeaders(req, auth, "key-unset-runtime-os-arch", false, nil, cfg)
 
-	assertClaudeFingerprint(t, req.Header, "claude-cli/2.1.60 (external, cli)", "0.70.0", "v22.0.0", mapStainlessOS(), mapStainlessArch())
+	assertClaudeFingerprint(t, req.Header, "claude-cli/2.1.60 (external, cli)", "0.70.0", "v22.0.0", helps.MapStainlessOS(), helps.MapStainlessArch())
 }
 
 func TestClaudeDeviceProfileStabilizationEnabled_DefaultFalse(t *testing.T) {
-	if claudeDeviceProfileStabilizationEnabled(nil) {
+	if helps.ClaudeDeviceProfileStabilizationEnabled(nil) {
 		t.Fatal("expected nil config to default to disabled stabilization")
 	}
-	if claudeDeviceProfileStabilizationEnabled(&config.Config{}) {
+	if helps.ClaudeDeviceProfileStabilizationEnabled(&config.Config{}) {
 		t.Fatal("expected unset stabilize-device-profile to default to disabled stabilization")
 	}
 }
@@ -796,8 +799,6 @@ func TestApplyClaudeToolPrefix_NestedToolReference(t *testing.T) {
 }
 
 func TestClaudeExecutor_ReusesUserIDAcrossModelsWhenCacheEnabled(t *testing.T) {
-	resetUserIDCache()
-
 	var userIDs []string
 	var requestModels []string
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
@@ -857,15 +858,13 @@ func TestClaudeExecutor_ReusesUserIDAcrossModelsWhenCacheEnabled(t *testing.T) {
 	if userIDs[0] != userIDs[1] {
 		t.Fatalf("expected user_id to be reused across models, got %q and %q", userIDs[0], userIDs[1])
 	}
-	if !isValidUserID(userIDs[0]) {
+	if !helps.IsValidUserID(userIDs[0]) {
 		t.Fatalf("user_id %q is not valid", userIDs[0])
 	}
 	t.Logf("✓ End-to-end test passed: Same user_id (%s) was used for both models", userIDs[0])
 }
 
 func TestClaudeExecutor_GeneratesNewUserIDByDefault(t *testing.T) {
-	resetUserIDCache()
-
 	var userIDs []string
 	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		body, _ := io.ReadAll(r.Body)
@@ -903,7 +902,7 @@ func TestClaudeExecutor_GeneratesNewUserIDByDefault(t *testing.T) {
 	if userIDs[0] == userIDs[1] {
 		t.Fatalf("expected user_id to change when caching is not enabled, got identical values %q", userIDs[0])
 	}
-	if !isValidUserID(userIDs[0]) || !isValidUserID(userIDs[1]) {
+	if !helps.IsValidUserID(userIDs[0]) || !helps.IsValidUserID(userIDs[1]) {
 		t.Fatalf("user_ids should be valid, got %q and %q", userIDs[0], userIDs[1])
 	}
 }
@@ -1183,6 +1182,83 @@ func testClaudeExecutorInvalidCompressedErrorBody(
 	}
 }
 
+func TestEnsureModelMaxTokens_UsesRegisteredMaxCompletionTokens(t *testing.T) {
+	reg := registry.GetGlobalRegistry()
+	clientID := "test-claude-max-completion-tokens-client"
+	modelID := "test-claude-max-completion-tokens-model"
+	reg.RegisterClient(clientID, "claude", []*registry.ModelInfo{{
+		ID:                  modelID,
+		Type:                "claude",
+		OwnedBy:             "anthropic",
+		Object:              "model",
+		Created:             time.Now().Unix(),
+		MaxCompletionTokens: 4096,
+		UserDefined:         true,
+	}})
+	defer reg.UnregisterClient(clientID)
+
+	input := []byte(`{"model":"test-claude-max-completion-tokens-model","messages":[{"role":"user","content":"hi"}]}`)
+	out := ensureModelMaxTokens(input, modelID)
+
+	if got := gjson.GetBytes(out, "max_tokens").Int(); got != 4096 {
+		t.Fatalf("max_tokens = %d, want %d", got, 4096)
+	}
+}
+
+func TestEnsureModelMaxTokens_DefaultsMissingValue(t *testing.T) {
+	reg := registry.GetGlobalRegistry()
+	clientID := "test-claude-default-max-tokens-client"
+	modelID := "test-claude-default-max-tokens-model"
+	reg.RegisterClient(clientID, "claude", []*registry.ModelInfo{{
+		ID:          modelID,
+		Type:        "claude",
+		OwnedBy:     "anthropic",
+		Object:      "model",
+		Created:     time.Now().Unix(),
+		UserDefined: true,
+	}})
+	defer reg.UnregisterClient(clientID)
+
+	input := []byte(`{"model":"test-claude-default-max-tokens-model","messages":[{"role":"user","content":"hi"}]}`)
+	out := ensureModelMaxTokens(input, modelID)
+
+	if got := gjson.GetBytes(out, "max_tokens").Int(); got != defaultModelMaxTokens {
+		t.Fatalf("max_tokens = %d, want %d", got, defaultModelMaxTokens)
+	}
+}
+
+func TestEnsureModelMaxTokens_PreservesExplicitValue(t *testing.T) {
+	reg := registry.GetGlobalRegistry()
+	clientID := "test-claude-preserve-max-tokens-client"
+	modelID := "test-claude-preserve-max-tokens-model"
+	reg.RegisterClient(clientID, "claude", []*registry.ModelInfo{{
+		ID:                  modelID,
+		Type:                "claude",
+		OwnedBy:             "anthropic",
+		Object:              "model",
+		Created:             time.Now().Unix(),
+		MaxCompletionTokens: 4096,
+		UserDefined:         true,
+	}})
+	defer reg.UnregisterClient(clientID)
+
+	input := []byte(`{"model":"test-claude-preserve-max-tokens-model","max_tokens":2048,"messages":[{"role":"user","content":"hi"}]}`)
+	out := ensureModelMaxTokens(input, modelID)
+
+	if got := gjson.GetBytes(out, "max_tokens").Int(); got != 2048 {
+		t.Fatalf("max_tokens = %d, want %d", got, 2048)
+	}
+}
+
+func TestEnsureModelMaxTokens_SkipsUnregisteredModel(t *testing.T) {
+	input := []byte(`{"model":"test-claude-unregistered-model","messages":[{"role":"user","content":"hi"}]}`)
+	out := ensureModelMaxTokens(input, "test-claude-unregistered-model")
+
+	if gjson.GetBytes(out, "max_tokens").Exists() {
+		t.Fatalf("max_tokens should remain unset, got %s", gjson.GetBytes(out, "max_tokens").Raw)
+	}
+}
+
 // TestClaudeExecutor_ExecuteStream_SetsIdentityAcceptEncoding verifies that streaming
 // requests use Accept-Encoding: identity so the upstream cannot respond with a
 // compressed SSE body that would silently break the line scanner.
@@ -1340,6 +1416,35 @@ func TestDecodeResponseBody_MagicByteGzipNoHeader(t *testing.T) {
 	}
 }
 
+// TestDecodeResponseBody_MagicByteZstdNoHeader verifies that decodeResponseBody
+// detects zstd-compressed content via magic bytes even when Content-Encoding is absent.
+func TestDecodeResponseBody_MagicByteZstdNoHeader(t *testing.T) {
+	const plaintext = "data: {\"type\":\"message_stop\"}\n"
+
+	var buf bytes.Buffer
+	enc, err := zstd.NewWriter(&buf)
+	if err != nil {
+		t.Fatalf("zstd.NewWriter: %v", err)
+	}
+	_, _ = enc.Write([]byte(plaintext))
+	_ = enc.Close()
+
+	rc := io.NopCloser(&buf)
+	decoded, err := decodeResponseBody(rc, "")
+	if err != nil {
+		t.Fatalf("decodeResponseBody error: %v", err)
+	}
+	defer decoded.Close()
+
+	got, err := io.ReadAll(decoded)
+	if err != nil {
+		t.Fatalf("ReadAll error: %v", err)
+	}
+	if string(got) != plaintext {
+		t.Errorf("decoded = %q, want %q", got, plaintext)
+	}
+}
+
 // TestDecodeResponseBody_PlainTextNoHeader verifies that decodeResponseBody returns
 // plain text untouched when Content-Encoding is absent and no magic bytes match.
 func TestDecodeResponseBody_PlainTextNoHeader(t *testing.T) {
@@ -1411,77 +1516,6 @@ func TestClaudeExecutor_ExecuteStream_GzipNoContentEncodingHeader(t *testing.T)
 	}
 }
 
-// TestClaudeExecutor_ExecuteStream_AcceptEncodingOverrideCannotBypassIdentity verifies
-// that injecting Accept-Encoding via auth.Attributes cannot override the stream
-// path's enforced identity encoding.
-func TestClaudeExecutor_ExecuteStream_AcceptEncodingOverrideCannotBypassIdentity(t *testing.T) {
-	var gotEncoding string
-	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		gotEncoding = r.Header.Get("Accept-Encoding")
-		w.Header().Set("Content-Type", "text/event-stream")
-		_, _ = w.Write([]byte("data: {\"type\":\"message_stop\"}\n\n"))
-	}))
-	defer server.Close()
-
-	executor := NewClaudeExecutor(&config.Config{})
-	// Inject Accept-Encoding via the custom header attribute mechanism.
-	auth := &cliproxyauth.Auth{Attributes: map[string]string{
-		"api_key":                "key-123",
-		"base_url":               server.URL,
-		"header:Accept-Encoding": "gzip, deflate, br, zstd",
-	}}
-	payload := []byte(`{"messages":[{"role":"user","content":[{"type":"text","text":"hi"}]}]}`)
-
-	result, err := executor.ExecuteStream(context.Background(), auth, cliproxyexecutor.Request{
-		Model:   "claude-3-5-sonnet-20241022",
-		Payload: payload,
-	}, cliproxyexecutor.Options{
-		SourceFormat: sdktranslator.FromString("claude"),
-	})
-	if err != nil {
-		t.Fatalf("ExecuteStream error: %v", err)
-	}
-	for chunk := range result.Chunks {
-		if chunk.Err != nil {
-			t.Fatalf("unexpected chunk error: %v", chunk.Err)
-		}
-	}
-
-	if gotEncoding != "identity" {
-		t.Errorf("Accept-Encoding = %q; stream path must enforce identity regardless of auth.Attributes override", gotEncoding)
-	}
-}
-
-// TestDecodeResponseBody_MagicByteZstdNoHeader verifies that decodeResponseBody
-// detects zstd-compressed content via magic bytes (28 b5 2f fd) even when
-// Content-Encoding is absent.
-func TestDecodeResponseBody_MagicByteZstdNoHeader(t *testing.T) {
-	const plaintext = "data: {\"type\":\"message_stop\"}\n"
-
-	var buf bytes.Buffer
-	enc, err := zstd.NewWriter(&buf)
-	if err != nil {
-		t.Fatalf("zstd.NewWriter: %v", err)
-	}
-	_, _ = enc.Write([]byte(plaintext))
-	_ = enc.Close()
-
-	rc := io.NopCloser(&buf)
-	decoded, err := decodeResponseBody(rc, "")
-	if err != nil {
-		t.Fatalf("decodeResponseBody error: %v", err)
-	}
-	defer decoded.Close()
-
-	got, err := io.ReadAll(decoded)
-	if err != nil {
-		t.Fatalf("ReadAll error: %v", err)
-	}
-	if string(got) != plaintext {
-		t.Errorf("decoded = %q, want %q", got, plaintext)
-	}
-}
-
 // TestClaudeExecutor_Execute_GzipErrorBodyNoContentEncodingHeader verifies that the
 // error path (4xx) correctly decompresses a gzip body even when the upstream omits
 // the Content-Encoding header.  This closes the gap left by PR #1771, which only
@@ -1565,6 +1599,45 @@ func TestClaudeExecutor_ExecuteStream_GzipErrorBodyNoContentEncodingHeader(t *te
 	}
 }
 
+// TestClaudeExecutor_ExecuteStream_AcceptEncodingOverrideCannotBypassIdentity verifies that the
+// streaming executor enforces Accept-Encoding: identity regardless of auth.Attributes override.
+func TestClaudeExecutor_ExecuteStream_AcceptEncodingOverrideCannotBypassIdentity(t *testing.T) {
+	var gotEncoding string
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		gotEncoding = r.Header.Get("Accept-Encoding")
+		w.Header().Set("Content-Type", "text/event-stream")
+		_, _ = w.Write([]byte("data: {\"type\":\"message_stop\"}\n\n"))
+	}))
+	defer server.Close()
+
+	executor := NewClaudeExecutor(&config.Config{})
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{
+		"api_key":                "key-123",
+		"base_url":               server.URL,
+		"header:Accept-Encoding": "gzip, deflate, br, zstd",
+	}}
+	payload := []byte(`{"messages":[{"role":"user","content":[{"type":"text","text":"hi"}]}]}`)
+
+	result, err := executor.ExecuteStream(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "claude-3-5-sonnet-20241022",
+		Payload: payload,
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("claude"),
+	})
+	if err != nil {
+		t.Fatalf("ExecuteStream error: %v", err)
+	}
+	for chunk := range result.Chunks {
+		if chunk.Err != nil {
+			t.Fatalf("unexpected chunk error: %v", chunk.Err)
+		}
+	}
+
+	if gotEncoding != "identity" {
+		t.Errorf("Accept-Encoding = %q; stream path must enforce identity regardless of auth.Attributes override", gotEncoding)
+	}
+}
+
 // Test case 1: String system prompt is preserved and converted to a content block
 func TestCheckSystemInstructionsWithMode_StringSystemPreserved(t *testing.T) {
 	payload := []byte(`{"system":"You are a helpful assistant.","messages":[{"role":"user","content":"hi"}]}`)
@@ -1648,3 +1721,155 @@ func TestCheckSystemInstructionsWithMode_StringWithSpecialChars(t *testing.T) {
 		t.Fatalf("blocks[2] text mangled, got %q", blocks[2].Get("text").String())
 	}
 }
+
+func TestClaudeExecutor_ExperimentalCCHSigningDisabledByDefaultKeepsLegacyHeader(t *testing.T) {
+	var seenBody []byte
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		seenBody = bytes.Clone(body)
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"id":"msg_1","type":"message","model":"claude-3-5-sonnet","role":"assistant","content":[{"type":"text","text":"ok"}],"usage":{"input_tokens":1,"output_tokens":1}}`))
+	}))
+	defer server.Close()
+
+	executor := NewClaudeExecutor(&config.Config{})
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{
+		"api_key":  "key-123",
+		"base_url": server.URL,
+	}}
+	payload := []byte(`{"messages":[{"role":"user","content":[{"type":"text","text":"hi"}]}]}`)
+
+	_, err := executor.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "claude-3-5-sonnet-20241022",
+		Payload: payload,
+	}, cliproxyexecutor.Options{SourceFormat: sdktranslator.FromString("claude")})
+	if err != nil {
+		t.Fatalf("Execute() error = %v", err)
+	}
+	if len(seenBody) == 0 {
+		t.Fatal("expected request body to be captured")
+	}
+
+	billingHeader := gjson.GetBytes(seenBody, "system.0.text").String()
+	if !strings.HasPrefix(billingHeader, "x-anthropic-billing-header:") {
+		t.Fatalf("system.0.text = %q, want billing header", billingHeader)
+	}
+	if strings.Contains(billingHeader, "cch=00000;") {
+		t.Fatalf("legacy mode should not forward cch placeholder, got %q", billingHeader)
+	}
+}
+
+func TestClaudeExecutor_ExperimentalCCHSigningOptInSignsFinalBody(t *testing.T) {
+	var seenBody []byte
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		body, _ := io.ReadAll(r.Body)
+		seenBody = bytes.Clone(body)
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"id":"msg_1","type":"message","model":"claude-3-5-sonnet","role":"assistant","content":[{"type":"text","text":"ok"}],"usage":{"input_tokens":1,"output_tokens":1}}`))
+	}))
+	defer server.Close()
+
+	executor := NewClaudeExecutor(&config.Config{
+		ClaudeKey: []config.ClaudeKey{{
+			APIKey:                 "key-123",
+			BaseURL:                server.URL,
+			ExperimentalCCHSigning: true,
+		}},
+	})
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{
+		"api_key":  "key-123",
+		"base_url": server.URL,
+	}}
+	const messageText = "please keep literal cch=00000 in this message"
+	payload := []byte(`{"messages":[{"role":"user","content":[{"type":"text","text":"please keep literal cch=00000 in this message"}]}]}`)
+
+	_, err := executor.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "claude-3-5-sonnet-20241022",
+		Payload: payload,
+	}, cliproxyexecutor.Options{SourceFormat: sdktranslator.FromString("claude")})
+	if err != nil {
+		t.Fatalf("Execute() error = %v", err)
+	}
+	if len(seenBody) == 0 {
+		t.Fatal("expected request body to be captured")
+	}
+	if got := gjson.GetBytes(seenBody, "messages.0.content.0.text").String(); got != messageText {
+		t.Fatalf("message text = %q, want %q", got, messageText)
+	}
+
+	billingPattern := regexp.MustCompile(`(x-anthropic-billing-header:[^"]*?\bcch=)([0-9a-f]{5})(;)`)
+	match := billingPattern.FindSubmatch(seenBody)
+	if match == nil {
+		t.Fatalf("expected signed billing header in body: %s", string(seenBody))
+	}
+	actualCCH := string(match[2])
+	unsignedBody := billingPattern.ReplaceAll(seenBody, []byte(`${1}00000${3}`))
+	wantCCH := fmt.Sprintf("%05x", xxHash64.Checksum(unsignedBody, 0x6E52736AC806831E)&0xFFFFF)
+	if actualCCH != wantCCH {
+		t.Fatalf("cch = %q, want %q\nbody: %s", actualCCH, wantCCH, string(seenBody))
+	}
+}
+
+func TestApplyCloaking_PreservesConfiguredStrictModeAndSensitiveWordsWhenModeOmitted(t *testing.T) {
+	cfg := &config.Config{
+		ClaudeKey: []config.ClaudeKey{{
+			APIKey: "key-123",
+			Cloak: &config.CloakConfig{
+				StrictMode:     true,
+				SensitiveWords: []string{"proxy"},
+			},
+		}},
+	}
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{"api_key": "key-123"}}
+	payload := []byte(`{"system":"proxy rules","messages":[{"role":"user","content":[{"type":"text","text":"proxy access"}]}]}`)
+
+	out := applyCloaking(context.Background(), cfg, auth, payload, "claude-3-5-sonnet-20241022", "key-123")
+
+	blocks := gjson.GetBytes(out, "system").Array()
+	if len(blocks) != 2 {
+		t.Fatalf("expected strict mode to keep only injected system blocks, got %d", len(blocks))
+	}
+	if got := gjson.GetBytes(out, "messages.0.content.0.text").String(); !strings.Contains(got, "\u200B") {
+		t.Fatalf("expected configured sensitive word obfuscation to apply, got %q", got)
+	}
+}
+
+func TestNormalizeClaudeTemperatureForThinking_AdaptiveCoercesToOne(t *testing.T) {
+	payload := []byte(`{"temperature":0,"thinking":{"type":"adaptive"},"output_config":{"effort":"max"}}`)
+	out := normalizeClaudeTemperatureForThinking(payload)
+
+	if got := gjson.GetBytes(out, "temperature").Float(); got != 1 {
+		t.Fatalf("temperature = %v, want 1", got)
+	}
+}
+
+func TestNormalizeClaudeTemperatureForThinking_EnabledCoercesToOne(t *testing.T) {
+	payload := []byte(`{"temperature":0.2,"thinking":{"type":"enabled","budget_tokens":2048}}`)
+	out := normalizeClaudeTemperatureForThinking(payload)
+
+	if got := gjson.GetBytes(out, "temperature").Float(); got != 1 {
+		t.Fatalf("temperature = %v, want 1", got)
+	}
+}
+
+func TestNormalizeClaudeTemperatureForThinking_NoThinkingLeavesTemperatureAlone(t *testing.T) {
+	payload := []byte(`{"temperature":0,"messages":[{"role":"user","content":"hi"}]}`)
+	out := normalizeClaudeTemperatureForThinking(payload)
+
+	if got := gjson.GetBytes(out, "temperature").Float(); got != 0 {
+		t.Fatalf("temperature = %v, want 0", got)
+	}
+}
+
+func TestNormalizeClaudeTemperatureForThinking_AfterForcedToolChoiceKeepsOriginalTemperature(t *testing.T) {
+	payload := []byte(`{"temperature":0,"thinking":{"type":"adaptive"},"output_config":{"effort":"max"},"tool_choice":{"type":"any"}}`)
+	out := disableThinkingIfToolChoiceForced(payload)
+	out = normalizeClaudeTemperatureForThinking(out)
+
+	if gjson.GetBytes(out, "thinking").Exists() {
+		t.Fatalf("thinking should be removed when tool_choice forces tool use")
+	}
+	if got := gjson.GetBytes(out, "temperature").Float(); got != 0 {
+		t.Fatalf("temperature = %v, want 0", got)
+	}
+}

internal/runtime/executor/claude_signing.go

@@ -0,0 +1,81 @@
+package executor
+
+import (
+	"fmt"
+	"regexp"
+	"strings"
+
+	xxHash64 "github.com/pierrec/xxHash/xxHash64"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	"github.com/tidwall/gjson"
+	"github.com/tidwall/sjson"
+)
+
+const claudeCCHSeed uint64 = 0x6E52736AC806831E
+
+var claudeBillingHeaderCCHPattern = regexp.MustCompile(`\bcch=([0-9a-f]{5});`)
+
+func signAnthropicMessagesBody(body []byte) []byte {
+	billingHeader := gjson.GetBytes(body, "system.0.text").String()
+	if !strings.HasPrefix(billingHeader, "x-anthropic-billing-header:") {
+		return body
+	}
+	if !claudeBillingHeaderCCHPattern.MatchString(billingHeader) {
+		return body
+	}
+
+	unsignedBillingHeader := claudeBillingHeaderCCHPattern.ReplaceAllString(billingHeader, "cch=00000;")
+	unsignedBody, err := sjson.SetBytes(body, "system.0.text", unsignedBillingHeader)
+	if err != nil {
+		return body
+	}
+
+	cch := fmt.Sprintf("%05x", xxHash64.Checksum(unsignedBody, claudeCCHSeed)&0xFFFFF)
+	signedBillingHeader := claudeBillingHeaderCCHPattern.ReplaceAllString(unsignedBillingHeader, "cch="+cch+";")
+	signedBody, err := sjson.SetBytes(unsignedBody, "system.0.text", signedBillingHeader)
+	if err != nil {
+		return unsignedBody
+	}
+	return signedBody
+}
+
+func resolveClaudeKeyConfig(cfg *config.Config, auth *cliproxyauth.Auth) *config.ClaudeKey {
+	if cfg == nil || auth == nil {
+		return nil
+	}
+
+	apiKey, baseURL := claudeCreds(auth)
+	if apiKey == "" {
+		return nil
+	}
+
+	for i := range cfg.ClaudeKey {
+		entry := &cfg.ClaudeKey[i]
+		cfgKey := strings.TrimSpace(entry.APIKey)
+		cfgBase := strings.TrimSpace(entry.BaseURL)
+		if !strings.EqualFold(cfgKey, apiKey) {
+			continue
+		}
+		if baseURL != "" && cfgBase != "" && !strings.EqualFold(cfgBase, baseURL) {
+			continue
+		}
+		return entry
+	}
+
+	return nil
+}
+
+// resolveClaudeKeyCloakConfig finds the matching ClaudeKey config and returns its CloakConfig.
+func resolveClaudeKeyCloakConfig(cfg *config.Config, auth *cliproxyauth.Auth) *config.CloakConfig {
+	entry := resolveClaudeKeyConfig(cfg, auth)
+	if entry == nil {
+		return nil
+	}
+	return entry.Cloak
+}
+
+func experimentalCCHSigningEnabled(cfg *config.Config, auth *cliproxyauth.Auth) bool {
+	entry := resolveClaudeKeyConfig(cfg, auth)
+	return entry != nil && entry.ExperimentalCCHSigning
+}

internal/runtime/executor/codex_executor.go

@@ -13,6 +13,7 @@ import (
 	codexauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/codex"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/misc"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -28,8 +29,8 @@ import (
 )
 
 const (
-	codexUserAgent  = "codex_cli_rs/0.116.0 (Mac OS 26.0.1; arm64) Apple_Terminal/464"
-	codexOriginator = "codex_cli_rs"
+	codexUserAgent  = "codex-tui/0.118.0 (Mac OS 26.3.1; arm64) iTerm.app/3.6.9 (codex-tui; 0.118.0)"
+	codexOriginator = "codex-tui"
 )
 
 var dataTag = []byte("data:")
@@ -73,7 +74,7 @@ func (e *CodexExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Auth
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -88,8 +89,8 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		baseURL = "https://chatgpt.com/backend-api/codex"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("codex")
@@ -106,16 +107,15 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 	body, _ = sjson.SetBytes(body, "stream", true)
 	body, _ = sjson.DeleteBytes(body, "previous_response_id")
 	body, _ = sjson.DeleteBytes(body, "prompt_cache_retention")
 	body, _ = sjson.DeleteBytes(body, "safety_identifier")
-	if !gjson.GetBytes(body, "instructions").Exists() {
-		body, _ = sjson.SetBytes(body, "instructions", "")
-	}
+	body, _ = sjson.DeleteBytes(body, "stream_options")
+	body = normalizeCodexInstructions(body)
 
 	url := strings.TrimSuffix(baseURL, "/") + "/responses"
 	httpReq, err := e.cacheHelper(ctx, from, url, req, body)
@@ -129,7 +129,7 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -140,10 +140,10 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		AuthType:  authType,
 		AuthValue: authValue,
 	})
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -151,20 +151,20 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 			log.Errorf("codex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = newCodexStatusErr(httpResp.StatusCode, b)
 		return resp, err
 	}
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 
 	lines := bytes.Split(data, []byte("\n"))
 	for _, line := range lines {
@@ -177,8 +177,8 @@ func (e *CodexExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 			continue
 		}
 
-		if detail, ok := parseCodexUsage(line); ok {
-			reporter.publish(ctx, detail)
+		if detail, ok := helps.ParseCodexUsage(line); ok {
+			reporter.Publish(ctx, detail)
 		}
 
 		var param any
@@ -198,8 +198,8 @@ func (e *CodexExecutor) executeCompact(ctx context.Context, auth *cliproxyauth.A
 		baseURL = "https://chatgpt.com/backend-api/codex"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("openai-response")
@@ -216,10 +216,11 @@ func (e *CodexExecutor) executeCompact(ctx context.Context, auth *cliproxyauth.A
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 	body, _ = sjson.DeleteBytes(body, "stream")
+	body = normalizeCodexInstructions(body)
 
 	url := strings.TrimSuffix(baseURL, "/") + "/responses/compact"
 	httpReq, err := e.cacheHelper(ctx, from, url, req, body)
@@ -233,7 +234,7 @@ func (e *CodexExecutor) executeCompact(ctx context.Context, auth *cliproxyauth.A
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -244,10 +245,10 @@ func (e *CodexExecutor) executeCompact(ctx context.Context, auth *cliproxyauth.A
 		AuthType:  authType,
 		AuthValue: authValue,
 	})
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -255,22 +256,22 @@ func (e *CodexExecutor) executeCompact(ctx context.Context, auth *cliproxyauth.A
 			log.Errorf("codex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = newCodexStatusErr(httpResp.StatusCode, b)
 		return resp, err
 	}
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseOpenAIUsage(data))
-	reporter.ensurePublished(ctx)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseOpenAIUsage(data))
+	reporter.EnsurePublished(ctx)
 	var param any
 	out := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, originalPayload, body, data, &param)
 	resp = cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}
@@ -288,8 +289,8 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		baseURL = "https://chatgpt.com/backend-api/codex"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("codex")
@@ -306,15 +307,14 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		return nil, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.DeleteBytes(body, "previous_response_id")
 	body, _ = sjson.DeleteBytes(body, "prompt_cache_retention")
 	body, _ = sjson.DeleteBytes(body, "safety_identifier")
+	body, _ = sjson.DeleteBytes(body, "stream_options")
 	body, _ = sjson.SetBytes(body, "model", baseModel)
-	if !gjson.GetBytes(body, "instructions").Exists() {
-		body, _ = sjson.SetBytes(body, "instructions", "")
-	}
+	body = normalizeCodexInstructions(body)
 
 	url := strings.TrimSuffix(baseURL, "/") + "/responses"
 	httpReq, err := e.cacheHelper(ctx, from, url, req, body)
@@ -328,7 +328,7 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -340,24 +340,24 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		data, readErr := io.ReadAll(httpResp.Body)
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("codex executor: close response body error: %v", errClose)
 		}
 		if readErr != nil {
-			recordAPIResponseError(ctx, e.cfg, readErr)
+			helps.RecordAPIResponseError(ctx, e.cfg, readErr)
 			return nil, readErr
 		}
-		appendAPIResponseChunk(ctx, e.cfg, data)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
 		err = newCodexStatusErr(httpResp.StatusCode, data)
 		return nil, err
 	}
@@ -374,13 +374,13 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
 
 			if bytes.HasPrefix(line, dataTag) {
 				data := bytes.TrimSpace(line[5:])
 				if gjson.GetBytes(data, "type").String() == "response.completed" {
-					if detail, ok := parseCodexUsage(data); ok {
-						reporter.publish(ctx, detail)
+					if detail, ok := helps.ParseCodexUsage(data); ok {
+						reporter.Publish(ctx, detail)
 					}
 				}
 			}
@@ -391,8 +391,8 @@ func (e *CodexExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 			}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -415,10 +415,9 @@ func (e *CodexExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Auth
 	body, _ = sjson.DeleteBytes(body, "previous_response_id")
 	body, _ = sjson.DeleteBytes(body, "prompt_cache_retention")
 	body, _ = sjson.DeleteBytes(body, "safety_identifier")
+	body, _ = sjson.DeleteBytes(body, "stream_options")
 	body, _ = sjson.SetBytes(body, "stream", false)
-	if !gjson.GetBytes(body, "instructions").Exists() {
-		body, _ = sjson.SetBytes(body, "instructions", "")
-	}
+	body = normalizeCodexInstructions(body)
 
 	enc, err := tokenizerForCodexModel(baseModel)
 	if err != nil {
@@ -597,18 +596,18 @@ func (e *CodexExecutor) Refresh(ctx context.Context, auth *cliproxyauth.Auth) (*
 }
 
 func (e *CodexExecutor) cacheHelper(ctx context.Context, from sdktranslator.Format, url string, req cliproxyexecutor.Request, rawJSON []byte) (*http.Request, error) {
-	var cache codexCache
+	var cache helps.CodexCache
 	if from == "claude" {
 		userIDResult := gjson.GetBytes(req.Payload, "metadata.user_id")
 		if userIDResult.Exists() {
 			key := fmt.Sprintf("%s-%s", req.Model, userIDResult.String())
 			var ok bool
-			if cache, ok = getCodexCache(key); !ok {
-				cache = codexCache{
+			if cache, ok = helps.GetCodexCache(key); !ok {
+				cache = helps.CodexCache{
 					ID:     uuid.New().String(),
 					Expire: time.Now().Add(1 * time.Hour),
 				}
-				setCodexCache(key, cache)
+				helps.SetCodexCache(key, cache)
 			}
 		}
 	} else if from == "openai-response" {
@@ -617,7 +616,7 @@ func (e *CodexExecutor) cacheHelper(ctx context.Context, from sdktranslator.Form
 			cache.ID = promptCacheKey.String()
 		}
 	} else if from == "openai" {
-		if apiKey := strings.TrimSpace(apiKeyFromContext(ctx)); apiKey != "" {
+		if apiKey := strings.TrimSpace(helps.APIKeyFromContext(ctx)); apiKey != "" {
 			cache.ID = uuid.NewSHA1(uuid.NameSpaceOID, []byte("cli-proxy-api:codex:prompt-cache:"+apiKey)).String()
 		}
 	}
@@ -630,7 +629,6 @@ func (e *CodexExecutor) cacheHelper(ctx context.Context, from sdktranslator.Form
 		return nil, err
 	}
 	if cache.ID != "" {
-		httpReq.Header.Set("Conversation_id", cache.ID)
 		httpReq.Header.Set("Session_id", cache.ID)
 	}
 	return httpReq, nil
@@ -645,13 +643,19 @@ func applyCodexHeaders(r *http.Request, auth *cliproxyauth.Auth, token string, s
 		ginHeaders = ginCtx.Request.Header
 	}
 
+	if ginHeaders.Get("X-Codex-Beta-Features") != "" {
+		r.Header.Set("X-Codex-Beta-Features", ginHeaders.Get("X-Codex-Beta-Features"))
+	}
 	misc.EnsureHeader(r.Header, ginHeaders, "Version", "")
-	misc.EnsureHeader(r.Header, ginHeaders, "Session_id", uuid.NewString())
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Codex-Turn-Metadata", "")
 	misc.EnsureHeader(r.Header, ginHeaders, "X-Client-Request-Id", "")
 	cfgUserAgent, _ := codexHeaderDefaults(cfg, auth)
 	ensureHeaderWithConfigPrecedence(r.Header, ginHeaders, "User-Agent", cfgUserAgent, codexUserAgent)
 
+	if strings.Contains(r.Header.Get("User-Agent"), "Mac OS") {
+		misc.EnsureHeader(r.Header, ginHeaders, "Session_id", uuid.NewString())
+	}
+
 	if stream {
 		r.Header.Set("Accept", "text/event-stream")
 	} else {
@@ -685,13 +689,47 @@ func applyCodexHeaders(r *http.Request, auth *cliproxyauth.Auth, token string, s
 }
 
 func newCodexStatusErr(statusCode int, body []byte) statusErr {
-	err := statusErr{code: statusCode, msg: string(body)}
-	if retryAfter := parseCodexRetryAfter(statusCode, body, time.Now()); retryAfter != nil {
+	errCode := statusCode
+	if isCodexModelCapacityError(body) {
+		errCode = http.StatusTooManyRequests
+	}
+	err := statusErr{code: errCode, msg: string(body)}
+	if retryAfter := parseCodexRetryAfter(errCode, body, time.Now()); retryAfter != nil {
 		err.retryAfter = retryAfter
 	}
 	return err
 }
 
+func normalizeCodexInstructions(body []byte) []byte {
+	instructions := gjson.GetBytes(body, "instructions")
+	if !instructions.Exists() || instructions.Type == gjson.Null {
+		body, _ = sjson.SetBytes(body, "instructions", "")
+	}
+	return body
+}
+
+func isCodexModelCapacityError(errorBody []byte) bool {
+	if len(errorBody) == 0 {
+		return false
+	}
+	candidates := []string{
+		gjson.GetBytes(errorBody, "error.message").String(),
+		gjson.GetBytes(errorBody, "message").String(),
+		string(errorBody),
+	}
+	for _, candidate := range candidates {
+		lower := strings.ToLower(strings.TrimSpace(candidate))
+		if lower == "" {
+			continue
+		}
+		if strings.Contains(lower, "selected model is at capacity") ||
+			strings.Contains(lower, "model is at capacity. please try a different model") {
+			return true
+		}
+	}
+	return false
+}
+
 func parseCodexRetryAfter(statusCode int, errorBody []byte, now time.Time) *time.Duration {
 	if statusCode != http.StatusTooManyRequests || len(errorBody) == 0 {
 		return nil

internal/runtime/executor/codex_executor_cache_test.go

@@ -42,8 +42,8 @@ func TestCodexExecutorCacheHelper_OpenAIChatCompletions_StablePromptCacheKeyFrom
 	if gotKey != expectedKey {
 		t.Fatalf("prompt_cache_key = %q, want %q", gotKey, expectedKey)
 	}
-	if gotConversation := httpReq.Header.Get("Conversation_id"); gotConversation != expectedKey {
-		t.Fatalf("Conversation_id = %q, want %q", gotConversation, expectedKey)
+	if gotConversation := httpReq.Header.Get("Conversation_id"); gotConversation != "" {
+		t.Fatalf("Conversation_id = %q, want empty", gotConversation)
 	}
 	if gotSession := httpReq.Header.Get("Session_id"); gotSession != expectedKey {
 		t.Fatalf("Session_id = %q, want %q", gotSession, expectedKey)

internal/runtime/executor/codex_executor_compact_test.go

@@ -0,0 +1,79 @@
+package executor
+
+import (
+	"context"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
+	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
+	"github.com/tidwall/gjson"
+)
+
+func TestCodexExecutorCompactAddsDefaultInstructions(t *testing.T) {
+	cases := []struct {
+		name    string
+		payload string
+	}{
+		{
+			name:    "missing instructions",
+			payload: `{"model":"gpt-5.4","input":"hello"}`,
+		},
+		{
+			name:    "null instructions",
+			payload: `{"model":"gpt-5.4","instructions":null,"input":"hello"}`,
+		},
+	}
+
+	for _, tc := range cases {
+		t.Run(tc.name, func(t *testing.T) {
+			var gotPath string
+			var gotBody []byte
+			server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+				gotPath = r.URL.Path
+				body, _ := io.ReadAll(r.Body)
+				gotBody = body
+				w.Header().Set("Content-Type", "application/json")
+				_, _ = w.Write([]byte(`{"id":"resp_1","object":"response.compaction","usage":{"input_tokens":1,"output_tokens":2,"total_tokens":3}}`))
+			}))
+			defer server.Close()
+
+			executor := NewCodexExecutor(&config.Config{})
+			auth := &cliproxyauth.Auth{Attributes: map[string]string{
+				"base_url": server.URL,
+				"api_key":  "test",
+			}}
+
+			resp, err := executor.Execute(context.Background(), auth, cliproxyexecutor.Request{
+				Model:   "gpt-5.4",
+				Payload: []byte(tc.payload),
+			}, cliproxyexecutor.Options{
+				SourceFormat: sdktranslator.FromString("openai-response"),
+				Alt:          "responses/compact",
+				Stream:       false,
+			})
+			if err != nil {
+				t.Fatalf("Execute error: %v", err)
+			}
+			if gotPath != "/responses/compact" {
+				t.Fatalf("path = %q, want %q", gotPath, "/responses/compact")
+			}
+			if !gjson.GetBytes(gotBody, "instructions").Exists() {
+				t.Fatalf("expected instructions in compact request body, got %s", string(gotBody))
+			}
+			if gjson.GetBytes(gotBody, "instructions").Type != gjson.String {
+				t.Fatalf("instructions type = %v, want string", gjson.GetBytes(gotBody, "instructions").Type)
+			}
+			if gjson.GetBytes(gotBody, "instructions").String() != "" {
+				t.Fatalf("instructions = %q, want empty string", gjson.GetBytes(gotBody, "instructions").String())
+			}
+			if string(resp.Payload) != `{"id":"resp_1","object":"response.compaction","usage":{"input_tokens":1,"output_tokens":2,"total_tokens":3}}` {
+				t.Fatalf("payload = %s", string(resp.Payload))
+			}
+		})
+	}
+}

internal/runtime/executor/codex_executor_instructions_test.go

@@ -0,0 +1,123 @@
+package executor
+
+import (
+	"context"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
+	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
+	"github.com/tidwall/gjson"
+)
+
+func TestCodexExecutorExecuteNormalizesNullInstructions(t *testing.T) {
+	var gotPath string
+	var gotBody []byte
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		gotPath = r.URL.Path
+		body, _ := io.ReadAll(r.Body)
+		gotBody = body
+		w.Header().Set("Content-Type", "text/event-stream")
+		_, _ = w.Write([]byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp_1\",\"object\":\"response\",\"created_at\":0,\"status\":\"completed\",\"background\":false,\"error\":null}}\n\n"))
+	}))
+	defer server.Close()
+
+	executor := NewCodexExecutor(&config.Config{})
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{
+		"base_url": server.URL,
+		"api_key":  "test",
+	}}
+
+	_, err := executor.Execute(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gpt-5.4",
+		Payload: []byte(`{"model":"gpt-5.4","instructions":null,"input":"hello"}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai-response"),
+		Stream:       false,
+	})
+	if err != nil {
+		t.Fatalf("Execute error: %v", err)
+	}
+	if gotPath != "/responses" {
+		t.Fatalf("path = %q, want %q", gotPath, "/responses")
+	}
+	if gjson.GetBytes(gotBody, "instructions").Type != gjson.String {
+		t.Fatalf("instructions type = %v, want string", gjson.GetBytes(gotBody, "instructions").Type)
+	}
+	if gjson.GetBytes(gotBody, "instructions").String() != "" {
+		t.Fatalf("instructions = %q, want empty string", gjson.GetBytes(gotBody, "instructions").String())
+	}
+}
+
+func TestCodexExecutorExecuteStreamNormalizesNullInstructions(t *testing.T) {
+	var gotPath string
+	var gotBody []byte
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		gotPath = r.URL.Path
+		body, _ := io.ReadAll(r.Body)
+		gotBody = body
+		w.Header().Set("Content-Type", "text/event-stream")
+		_, _ = w.Write([]byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp_1\",\"object\":\"response\",\"created_at\":0,\"status\":\"completed\",\"background\":false,\"error\":null}}\n\n"))
+	}))
+	defer server.Close()
+
+	executor := NewCodexExecutor(&config.Config{})
+	auth := &cliproxyauth.Auth{Attributes: map[string]string{
+		"base_url": server.URL,
+		"api_key":  "test",
+	}}
+
+	result, err := executor.ExecuteStream(context.Background(), auth, cliproxyexecutor.Request{
+		Model:   "gpt-5.4",
+		Payload: []byte(`{"model":"gpt-5.4","instructions":null,"input":"hello"}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai-response"),
+		Stream:       true,
+	})
+	if err != nil {
+		t.Fatalf("ExecuteStream error: %v", err)
+	}
+	for range result.Chunks {
+	}
+	if gotPath != "/responses" {
+		t.Fatalf("path = %q, want %q", gotPath, "/responses")
+	}
+	if gjson.GetBytes(gotBody, "instructions").Type != gjson.String {
+		t.Fatalf("instructions type = %v, want string", gjson.GetBytes(gotBody, "instructions").Type)
+	}
+	if gjson.GetBytes(gotBody, "instructions").String() != "" {
+		t.Fatalf("instructions = %q, want empty string", gjson.GetBytes(gotBody, "instructions").String())
+	}
+}
+
+func TestCodexExecutorCountTokensTreatsNullInstructionsAsEmpty(t *testing.T) {
+	executor := NewCodexExecutor(&config.Config{})
+
+	nullResp, err := executor.CountTokens(context.Background(), nil, cliproxyexecutor.Request{
+		Model:   "gpt-5.4",
+		Payload: []byte(`{"model":"gpt-5.4","instructions":null,"input":"hello"}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai-response"),
+	})
+	if err != nil {
+		t.Fatalf("CountTokens(null) error: %v", err)
+	}
+
+	emptyResp, err := executor.CountTokens(context.Background(), nil, cliproxyexecutor.Request{
+		Model:   "gpt-5.4",
+		Payload: []byte(`{"model":"gpt-5.4","instructions":"","input":"hello"}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai-response"),
+	})
+	if err != nil {
+		t.Fatalf("CountTokens(empty) error: %v", err)
+	}
+
+	if string(nullResp.Payload) != string(emptyResp.Payload) {
+		t.Fatalf("token count payload mismatch:\nnull=%s\nempty=%s", string(nullResp.Payload), string(emptyResp.Payload))
+	}
+}

internal/runtime/executor/codex_executor_retry_test.go

@@ -60,6 +60,19 @@ func TestParseCodexRetryAfter(t *testing.T) {
 	})
 }
 
+func TestNewCodexStatusErrTreatsCapacityAsRetryableRateLimit(t *testing.T) {
+	body := []byte(`{"error":{"message":"Selected model is at capacity. Please try a different model."}}`)
+
+	err := newCodexStatusErr(http.StatusBadRequest, body)
+
+	if got := err.StatusCode(); got != http.StatusTooManyRequests {
+		t.Fatalf("status code = %d, want %d", got, http.StatusTooManyRequests)
+	}
+	if err.RetryAfter() != nil {
+		t.Fatalf("expected nil explicit retryAfter for capacity fallback, got %v", *err.RetryAfter())
+	}
+}
+
 func itoa(v int64) string {
 	return strconv.FormatInt(v, 10)
 }

internal/runtime/executor/codex_websockets_executor.go

@@ -15,10 +15,12 @@ import (
 	"sync"
 	"time"
 
+	"github.com/gin-gonic/gin"
 	"github.com/google/uuid"
 	"github.com/gorilla/websocket"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/misc"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -44,10 +46,18 @@ const (
 type CodexWebsocketsExecutor struct {
 	*CodexExecutor
 
-	sessMu   sync.Mutex
+	store *codexWebsocketSessionStore
+}
+
+type codexWebsocketSessionStore struct {
+	mu       sync.Mutex
 	sessions map[string]*codexWebsocketSession
 }
 
+var globalCodexWebsocketSessionStore = &codexWebsocketSessionStore{
+	sessions: make(map[string]*codexWebsocketSession),
+}
+
 type codexWebsocketSession struct {
 	sessionID string
 
@@ -71,7 +81,7 @@ type codexWebsocketSession struct {
 func NewCodexWebsocketsExecutor(cfg *config.Config) *CodexWebsocketsExecutor {
 	return &CodexWebsocketsExecutor{
 		CodexExecutor: NewCodexExecutor(cfg),
-		sessions:      make(map[string]*codexWebsocketSession),
+		store:         globalCodexWebsocketSessionStore,
 	}
 }
 
@@ -155,8 +165,8 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		baseURL = "https://chatgpt.com/backend-api/codex"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("codex")
@@ -173,8 +183,8 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 	body, _ = sjson.SetBytes(body, "stream", true)
 	body, _ = sjson.DeleteBytes(body, "previous_response_id")
@@ -209,7 +219,7 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 	}
 
 	wsReqBody := buildCodexWebsocketRequestBody(body)
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	wsReqLog := helps.UpstreamRequestLog{
 		URL:       wsURL,
 		Method:    "WEBSOCKET",
 		Headers:   wsHeaders.Clone(),
@@ -219,16 +229,14 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		AuthLabel: authLabel,
 		AuthType:  authType,
 		AuthValue: authValue,
-	})
+	}
+	helps.RecordAPIWebsocketRequest(ctx, e.cfg, wsReqLog)
 
 	conn, respHS, errDial := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
-	if respHS != nil {
-		recordAPIResponseMetadata(ctx, e.cfg, respHS.StatusCode, respHS.Header.Clone())
-	}
 	if errDial != nil {
 		bodyErr := websocketHandshakeBody(respHS)
-		if len(bodyErr) > 0 {
-			appendAPIResponseChunk(ctx, e.cfg, bodyErr)
+		if respHS != nil {
+			helps.RecordAPIWebsocketUpgradeRejection(ctx, e.cfg, websocketUpgradeRequestLog(wsReqLog), respHS.StatusCode, respHS.Header.Clone(), bodyErr)
 		}
 		if respHS != nil && respHS.StatusCode == http.StatusUpgradeRequired {
 			return e.CodexExecutor.Execute(ctx, auth, req, opts)
@@ -236,10 +244,10 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		if respHS != nil && respHS.StatusCode > 0 {
 			return resp, statusErr{code: respHS.StatusCode, msg: string(bodyErr)}
 		}
-		recordAPIResponseError(ctx, e.cfg, errDial)
+		helps.RecordAPIWebsocketError(ctx, e.cfg, "dial", errDial)
 		return resp, errDial
 	}
-	closeHTTPResponseBody(respHS, "codex websockets executor: close handshake response body error")
+	recordAPIWebsocketHandshake(ctx, e.cfg, respHS)
 	if sess == nil {
 		logCodexWebsocketConnected(executionSessionID, authID, wsURL)
 		defer func() {
@@ -268,10 +276,10 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 			// Retry once with a fresh websocket connection. This is mainly to handle
 			// upstream closing the socket between sequential requests within the same
 			// execution session.
-			connRetry, _, errDialRetry := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
+			connRetry, respHSRetry, errDialRetry := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
 			if errDialRetry == nil && connRetry != nil {
 				wsReqBodyRetry := buildCodexWebsocketRequestBody(body)
-				recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+				helps.RecordAPIWebsocketRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 					URL:       wsURL,
 					Method:    "WEBSOCKET",
 					Headers:   wsHeaders.Clone(),
@@ -282,20 +290,22 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 					AuthType:  authType,
 					AuthValue: authValue,
 				})
+				recordAPIWebsocketHandshake(ctx, e.cfg, respHSRetry)
 				if errSendRetry := writeCodexWebsocketMessage(sess, connRetry, wsReqBodyRetry); errSendRetry == nil {
 					conn = connRetry
 					wsReqBody = wsReqBodyRetry
 				} else {
 					e.invalidateUpstreamConn(sess, connRetry, "send_error", errSendRetry)
-					recordAPIResponseError(ctx, e.cfg, errSendRetry)
+					helps.RecordAPIWebsocketError(ctx, e.cfg, "send_retry", errSendRetry)
 					return resp, errSendRetry
 				}
 			} else {
-				recordAPIResponseError(ctx, e.cfg, errDialRetry)
+				closeHTTPResponseBody(respHSRetry, "codex websockets executor: close handshake response body error")
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "dial_retry", errDialRetry)
 				return resp, errDialRetry
 			}
 		} else {
-			recordAPIResponseError(ctx, e.cfg, errSend)
+			helps.RecordAPIWebsocketError(ctx, e.cfg, "send", errSend)
 			return resp, errSend
 		}
 	}
@@ -306,7 +316,7 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		}
 		msgType, payload, errRead := readCodexWebsocketMessage(ctx, sess, conn, readCh)
 		if errRead != nil {
-			recordAPIResponseError(ctx, e.cfg, errRead)
+			helps.RecordAPIWebsocketError(ctx, e.cfg, "read", errRead)
 			return resp, errRead
 		}
 		if msgType != websocket.TextMessage {
@@ -315,7 +325,7 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 				if sess != nil {
 					e.invalidateUpstreamConn(sess, conn, "unexpected_binary", err)
 				}
-				recordAPIResponseError(ctx, e.cfg, err)
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "unexpected_binary", err)
 				return resp, err
 			}
 			continue
@@ -325,21 +335,21 @@ func (e *CodexWebsocketsExecutor) Execute(ctx context.Context, auth *cliproxyaut
 		if len(payload) == 0 {
 			continue
 		}
-		appendAPIResponseChunk(ctx, e.cfg, payload)
+		helps.AppendAPIWebsocketResponse(ctx, e.cfg, payload)
 
 		if wsErr, ok := parseCodexWebsocketError(payload); ok {
 			if sess != nil {
 				e.invalidateUpstreamConn(sess, conn, "upstream_error", wsErr)
 			}
-			recordAPIResponseError(ctx, e.cfg, wsErr)
+			helps.RecordAPIWebsocketError(ctx, e.cfg, "upstream_error", wsErr)
 			return resp, wsErr
 		}
 
 		payload = normalizeCodexWebsocketCompletion(payload)
 		eventType := gjson.GetBytes(payload, "type").String()
 		if eventType == "response.completed" {
-			if detail, ok := parseCodexUsage(payload); ok {
-				reporter.publish(ctx, detail)
+			if detail, ok := helps.ParseCodexUsage(payload); ok {
+				reporter.Publish(ctx, detail)
 			}
 			var param any
 			out := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, originalPayload, body, payload, &param)
@@ -364,8 +374,8 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 		baseURL = "https://chatgpt.com/backend-api/codex"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("codex")
@@ -376,8 +386,8 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 		return nil, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, body, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, body, requestedModel)
 
 	httpURL := strings.TrimSuffix(baseURL, "/") + "/responses"
 	wsURL, err := buildCodexResponsesWebsocketURL(httpURL)
@@ -403,7 +413,7 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 	}
 
 	wsReqBody := buildCodexWebsocketRequestBody(body)
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	wsReqLog := helps.UpstreamRequestLog{
 		URL:       wsURL,
 		Method:    "WEBSOCKET",
 		Headers:   wsHeaders.Clone(),
@@ -413,18 +423,18 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 		AuthLabel: authLabel,
 		AuthType:  authType,
 		AuthValue: authValue,
-	})
+	}
+	helps.RecordAPIWebsocketRequest(ctx, e.cfg, wsReqLog)
 
 	conn, respHS, errDial := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
 	var upstreamHeaders http.Header
 	if respHS != nil {
 		upstreamHeaders = respHS.Header.Clone()
-		recordAPIResponseMetadata(ctx, e.cfg, respHS.StatusCode, respHS.Header.Clone())
 	}
 	if errDial != nil {
 		bodyErr := websocketHandshakeBody(respHS)
-		if len(bodyErr) > 0 {
-			appendAPIResponseChunk(ctx, e.cfg, bodyErr)
+		if respHS != nil {
+			helps.RecordAPIWebsocketUpgradeRejection(ctx, e.cfg, websocketUpgradeRequestLog(wsReqLog), respHS.StatusCode, respHS.Header.Clone(), bodyErr)
 		}
 		if respHS != nil && respHS.StatusCode == http.StatusUpgradeRequired {
 			return e.CodexExecutor.ExecuteStream(ctx, auth, req, opts)
@@ -432,13 +442,13 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 		if respHS != nil && respHS.StatusCode > 0 {
 			return nil, statusErr{code: respHS.StatusCode, msg: string(bodyErr)}
 		}
-		recordAPIResponseError(ctx, e.cfg, errDial)
+		helps.RecordAPIWebsocketError(ctx, e.cfg, "dial", errDial)
 		if sess != nil {
 			sess.reqMu.Unlock()
 		}
 		return nil, errDial
 	}
-	closeHTTPResponseBody(respHS, "codex websockets executor: close handshake response body error")
+	recordAPIWebsocketHandshake(ctx, e.cfg, respHS)
 
 	if sess == nil {
 		logCodexWebsocketConnected(executionSessionID, authID, wsURL)
@@ -451,20 +461,21 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 	}
 
 	if errSend := writeCodexWebsocketMessage(sess, conn, wsReqBody); errSend != nil {
-		recordAPIResponseError(ctx, e.cfg, errSend)
+		helps.RecordAPIWebsocketError(ctx, e.cfg, "send", errSend)
 		if sess != nil {
 			e.invalidateUpstreamConn(sess, conn, "send_error", errSend)
 
 			// Retry once with a new websocket connection for the same execution session.
-			connRetry, _, errDialRetry := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
+			connRetry, respHSRetry, errDialRetry := e.ensureUpstreamConn(ctx, auth, sess, authID, wsURL, wsHeaders)
 			if errDialRetry != nil || connRetry == nil {
-				recordAPIResponseError(ctx, e.cfg, errDialRetry)
+				closeHTTPResponseBody(respHSRetry, "codex websockets executor: close handshake response body error")
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "dial_retry", errDialRetry)
 				sess.clearActive(readCh)
 				sess.reqMu.Unlock()
 				return nil, errDialRetry
 			}
 			wsReqBodyRetry := buildCodexWebsocketRequestBody(body)
-			recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+			helps.RecordAPIWebsocketRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 				URL:       wsURL,
 				Method:    "WEBSOCKET",
 				Headers:   wsHeaders.Clone(),
@@ -475,8 +486,9 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 				AuthType:  authType,
 				AuthValue: authValue,
 			})
+			recordAPIWebsocketHandshake(ctx, e.cfg, respHSRetry)
 			if errSendRetry := writeCodexWebsocketMessage(sess, connRetry, wsReqBodyRetry); errSendRetry != nil {
-				recordAPIResponseError(ctx, e.cfg, errSendRetry)
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "send_retry", errSendRetry)
 				e.invalidateUpstreamConn(sess, connRetry, "send_error", errSendRetry)
 				sess.clearActive(readCh)
 				sess.reqMu.Unlock()
@@ -542,8 +554,8 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 				}
 				terminateReason = "read_error"
 				terminateErr = errRead
-				recordAPIResponseError(ctx, e.cfg, errRead)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "read", errRead)
+				reporter.PublishFailure(ctx)
 				_ = send(cliproxyexecutor.StreamChunk{Err: errRead})
 				return
 			}
@@ -552,8 +564,8 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 					err = fmt.Errorf("codex websockets executor: unexpected binary message")
 					terminateReason = "unexpected_binary"
 					terminateErr = err
-					recordAPIResponseError(ctx, e.cfg, err)
-					reporter.publishFailure(ctx)
+					helps.RecordAPIWebsocketError(ctx, e.cfg, "unexpected_binary", err)
+					reporter.PublishFailure(ctx)
 					if sess != nil {
 						e.invalidateUpstreamConn(sess, conn, "unexpected_binary", err)
 					}
@@ -567,13 +579,13 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 			if len(payload) == 0 {
 				continue
 			}
-			appendAPIResponseChunk(ctx, e.cfg, payload)
+			helps.AppendAPIWebsocketResponse(ctx, e.cfg, payload)
 
 			if wsErr, ok := parseCodexWebsocketError(payload); ok {
 				terminateReason = "upstream_error"
 				terminateErr = wsErr
-				recordAPIResponseError(ctx, e.cfg, wsErr)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIWebsocketError(ctx, e.cfg, "upstream_error", wsErr)
+				reporter.PublishFailure(ctx)
 				if sess != nil {
 					e.invalidateUpstreamConn(sess, conn, "upstream_error", wsErr)
 				}
@@ -584,8 +596,8 @@ func (e *CodexWebsocketsExecutor) ExecuteStream(ctx context.Context, auth *clipr
 			payload = normalizeCodexWebsocketCompletion(payload)
 			eventType := gjson.GetBytes(payload, "type").String()
 			if eventType == "response.completed" || eventType == "response.done" {
-				if detail, ok := parseCodexUsage(payload); ok {
-					reporter.publish(ctx, detail)
+				if detail, ok := helps.ParseCodexUsage(payload); ok {
+					reporter.Publish(ctx, detail)
 				}
 			}
 
@@ -767,19 +779,19 @@ func applyCodexPromptCacheHeaders(from sdktranslator.Format, req cliproxyexecuto
 		return rawJSON, headers
 	}
 
-	var cache codexCache
+	var cache helps.CodexCache
 	if from == "claude" {
 		userIDResult := gjson.GetBytes(req.Payload, "metadata.user_id")
 		if userIDResult.Exists() {
 			key := fmt.Sprintf("%s-%s", req.Model, userIDResult.String())
-			if cached, ok := getCodexCache(key); ok {
+			if cached, ok := helps.GetCodexCache(key); ok {
 				cache = cached
 			} else {
-				cache = codexCache{
+				cache = helps.CodexCache{
 					ID:     uuid.New().String(),
 					Expire: time.Now().Add(1 * time.Hour),
 				}
-				setCodexCache(key, cache)
+				helps.SetCodexCache(key, cache)
 			}
 		}
 	} else if from == "openai-response" {
@@ -791,7 +803,6 @@ func applyCodexPromptCacheHeaders(from sdktranslator.Format, req cliproxyexecuto
 	if cache.ID != "" {
 		rawJSON, _ = sjson.SetBytes(rawJSON, "prompt_cache_key", cache.ID)
 		headers.Set("Conversation_id", cache.ID)
-		headers.Set("Session_id", cache.ID)
 	}
 
 	return rawJSON, headers
@@ -806,11 +817,11 @@ func applyCodexWebsocketHeaders(ctx context.Context, headers http.Header, auth *
 	}
 
 	var ginHeaders http.Header
-	if ginCtx := ginContextFrom(ctx); ginCtx != nil && ginCtx.Request != nil {
-		ginHeaders = ginCtx.Request.Header
+	if ginCtx, ok := ctx.Value("gin").(*gin.Context); ok && ginCtx != nil && ginCtx.Request != nil {
+		ginHeaders = ginCtx.Request.Header.Clone()
 	}
 
-	cfgUserAgent, cfgBetaFeatures := codexHeaderDefaults(cfg, auth)
+	_, cfgBetaFeatures := codexHeaderDefaults(cfg, auth)
 	ensureHeaderWithPriority(headers, ginHeaders, "x-codex-beta-features", cfgBetaFeatures, "")
 	misc.EnsureHeader(headers, ginHeaders, "x-codex-turn-state", "")
 	misc.EnsureHeader(headers, ginHeaders, "x-codex-turn-metadata", "")
@@ -826,8 +837,10 @@ func applyCodexWebsocketHeaders(ctx context.Context, headers http.Header, auth *
 		betaHeader = codexResponsesWebsocketBetaHeaderValue
 	}
 	headers.Set("OpenAI-Beta", betaHeader)
-	misc.EnsureHeader(headers, ginHeaders, "Session_id", uuid.NewString())
-	ensureHeaderWithConfigPrecedence(headers, ginHeaders, "User-Agent", cfgUserAgent, codexUserAgent)
+	if strings.Contains(headers.Get("User-Agent"), "Mac OS") {
+		misc.EnsureHeader(headers, ginHeaders, "Session_id", uuid.NewString())
+	}
+	headers.Del("User-Agent")
 
 	isAPIKey := false
 	if auth != nil && auth.Attributes != nil {
@@ -1011,6 +1024,32 @@ func encodeCodexWebsocketAsSSE(payload []byte) []byte {
 	return line
 }
 
+func websocketUpgradeRequestLog(info helps.UpstreamRequestLog) helps.UpstreamRequestLog {
+	upgradeInfo := info
+	upgradeInfo.URL = helps.WebsocketUpgradeRequestURL(info.URL)
+	upgradeInfo.Method = http.MethodGet
+	upgradeInfo.Body = nil
+	upgradeInfo.Headers = info.Headers.Clone()
+	if upgradeInfo.Headers == nil {
+		upgradeInfo.Headers = make(http.Header)
+	}
+	if strings.TrimSpace(upgradeInfo.Headers.Get("Connection")) == "" {
+		upgradeInfo.Headers.Set("Connection", "Upgrade")
+	}
+	if strings.TrimSpace(upgradeInfo.Headers.Get("Upgrade")) == "" {
+		upgradeInfo.Headers.Set("Upgrade", "websocket")
+	}
+	return upgradeInfo
+}
+
+func recordAPIWebsocketHandshake(ctx context.Context, cfg *config.Config, resp *http.Response) {
+	if resp == nil {
+		return
+	}
+	helps.RecordAPIWebsocketHandshake(ctx, cfg, resp.StatusCode, resp.Header.Clone())
+	closeHTTPResponseBody(resp, "codex websockets executor: close handshake response body error")
+}
+
 func websocketHandshakeBody(resp *http.Response) []byte {
 	if resp == nil || resp.Body == nil {
 		return nil
@@ -1055,16 +1094,23 @@ func (e *CodexWebsocketsExecutor) getOrCreateSession(sessionID string) *codexWeb
 	if sessionID == "" {
 		return nil
 	}
-	e.sessMu.Lock()
-	defer e.sessMu.Unlock()
-	if e.sessions == nil {
-		e.sessions = make(map[string]*codexWebsocketSession)
+	if e == nil {
+		return nil
 	}
-	if sess, ok := e.sessions[sessionID]; ok && sess != nil {
+	store := e.store
+	if store == nil {
+		store = globalCodexWebsocketSessionStore
+	}
+	store.mu.Lock()
+	defer store.mu.Unlock()
+	if store.sessions == nil {
+		store.sessions = make(map[string]*codexWebsocketSession)
+	}
+	if sess, ok := store.sessions[sessionID]; ok && sess != nil {
 		return sess
 	}
 	sess := &codexWebsocketSession{sessionID: sessionID}
-	e.sessions[sessionID] = sess
+	store.sessions[sessionID] = sess
 	return sess
 }
 
@@ -1210,14 +1256,20 @@ func (e *CodexWebsocketsExecutor) CloseExecutionSession(sessionID string) {
 		return
 	}
 	if sessionID == cliproxyauth.CloseAllExecutionSessionsID {
-		e.closeAllExecutionSessions("executor_replaced")
+		// Executor replacement can happen during hot reload (config/credential changes).
+		// Do not force-close upstream websocket sessions here, otherwise in-flight
+		// downstream websocket requests get interrupted.
 		return
 	}
 
-	e.sessMu.Lock()
-	sess := e.sessions[sessionID]
-	delete(e.sessions, sessionID)
-	e.sessMu.Unlock()
+	store := e.store
+	if store == nil {
+		store = globalCodexWebsocketSessionStore
+	}
+	store.mu.Lock()
+	sess := store.sessions[sessionID]
+	delete(store.sessions, sessionID)
+	store.mu.Unlock()
 
 	e.closeExecutionSession(sess, "session_closed")
 }
@@ -1227,15 +1279,19 @@ func (e *CodexWebsocketsExecutor) closeAllExecutionSessions(reason string) {
 		return
 	}
 
-	e.sessMu.Lock()
-	sessions := make([]*codexWebsocketSession, 0, len(e.sessions))
-	for sessionID, sess := range e.sessions {
-		delete(e.sessions, sessionID)
+	store := e.store
+	if store == nil {
+		store = globalCodexWebsocketSessionStore
+	}
+	store.mu.Lock()
+	sessions := make([]*codexWebsocketSession, 0, len(store.sessions))
+	for sessionID, sess := range store.sessions {
+		delete(store.sessions, sessionID)
 		if sess != nil {
 			sessions = append(sessions, sess)
 		}
 	}
-	e.sessMu.Unlock()
+	store.mu.Unlock()
 
 	for i := range sessions {
 		e.closeExecutionSession(sessions[i], reason)
@@ -1243,6 +1299,10 @@ func (e *CodexWebsocketsExecutor) closeAllExecutionSessions(reason string) {
 }
 
 func (e *CodexWebsocketsExecutor) closeExecutionSession(sess *codexWebsocketSession, reason string) {
+	closeCodexWebsocketSession(sess, reason)
+}
+
+func closeCodexWebsocketSession(sess *codexWebsocketSession, reason string) {
 	if sess == nil {
 		return
 	}
@@ -1283,6 +1343,69 @@ func logCodexWebsocketDisconnected(sessionID string, authID string, wsURL string
 	log.Infof("codex websockets: upstream disconnected session=%s auth=%s url=%s reason=%s", strings.TrimSpace(sessionID), strings.TrimSpace(authID), strings.TrimSpace(wsURL), strings.TrimSpace(reason))
 }
 
+// CloseCodexWebsocketSessionsForAuthID closes all active Codex upstream websocket sessions
+// associated with the supplied auth ID.
+func CloseCodexWebsocketSessionsForAuthID(authID string, reason string) {
+	authID = strings.TrimSpace(authID)
+	if authID == "" {
+		return
+	}
+	reason = strings.TrimSpace(reason)
+	if reason == "" {
+		reason = "auth_removed"
+	}
+
+	store := globalCodexWebsocketSessionStore
+	if store == nil {
+		return
+	}
+
+	type sessionItem struct {
+		sessionID string
+		sess      *codexWebsocketSession
+	}
+
+	store.mu.Lock()
+	items := make([]sessionItem, 0, len(store.sessions))
+	for sessionID, sess := range store.sessions {
+		items = append(items, sessionItem{sessionID: sessionID, sess: sess})
+	}
+	store.mu.Unlock()
+
+	matches := make([]sessionItem, 0)
+	for i := range items {
+		sess := items[i].sess
+		if sess == nil {
+			continue
+		}
+		sess.connMu.Lock()
+		sessAuthID := strings.TrimSpace(sess.authID)
+		sess.connMu.Unlock()
+		if sessAuthID == authID {
+			matches = append(matches, items[i])
+		}
+	}
+	if len(matches) == 0 {
+		return
+	}
+
+	toClose := make([]*codexWebsocketSession, 0, len(matches))
+	store.mu.Lock()
+	for i := range matches {
+		current, ok := store.sessions[matches[i].sessionID]
+		if !ok || current == nil || current != matches[i].sess {
+			continue
+		}
+		delete(store.sessions, matches[i].sessionID)
+		toClose = append(toClose, current)
+	}
+	store.mu.Unlock()
+
+	for i := range toClose {
+		closeCodexWebsocketSession(toClose[i], reason)
+	}
+}
+
 // CodexAutoExecutor routes Codex requests to the websocket transport only when:
 //  1. The downstream transport is websocket, and
 //  2. The selected auth enables websockets.

internal/runtime/executor/codex_websockets_executor_store_test.go

@@ -0,0 +1,48 @@
+package executor
+
+import (
+	"testing"
+
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+)
+
+func TestCodexWebsocketsExecutor_SessionStoreSurvivesExecutorReplacement(t *testing.T) {
+	sessionID := "test-session-store-survives-replace"
+
+	globalCodexWebsocketSessionStore.mu.Lock()
+	delete(globalCodexWebsocketSessionStore.sessions, sessionID)
+	globalCodexWebsocketSessionStore.mu.Unlock()
+
+	exec1 := NewCodexWebsocketsExecutor(nil)
+	sess1 := exec1.getOrCreateSession(sessionID)
+	if sess1 == nil {
+		t.Fatalf("expected session to be created")
+	}
+
+	exec2 := NewCodexWebsocketsExecutor(nil)
+	sess2 := exec2.getOrCreateSession(sessionID)
+	if sess2 == nil {
+		t.Fatalf("expected session to be available across executors")
+	}
+	if sess1 != sess2 {
+		t.Fatalf("expected the same session instance across executors")
+	}
+
+	exec1.CloseExecutionSession(cliproxyauth.CloseAllExecutionSessionsID)
+
+	globalCodexWebsocketSessionStore.mu.Lock()
+	_, stillPresent := globalCodexWebsocketSessionStore.sessions[sessionID]
+	globalCodexWebsocketSessionStore.mu.Unlock()
+	if !stillPresent {
+		t.Fatalf("expected session to remain after executor replacement close marker")
+	}
+
+	exec2.CloseExecutionSession(sessionID)
+
+	globalCodexWebsocketSessionStore.mu.Lock()
+	_, presentAfterClose := globalCodexWebsocketSessionStore.sessions[sessionID]
+	globalCodexWebsocketSessionStore.mu.Unlock()
+	if presentAfterClose {
+		t.Fatalf("expected session to be removed after explicit close")
+	}
+}

internal/runtime/executor/codex_websockets_executor_test.go

@@ -38,8 +38,8 @@ func TestApplyCodexWebsocketHeadersDefaultsToCurrentResponsesBeta(t *testing.T)
 	if got := headers.Get("OpenAI-Beta"); got != codexResponsesWebsocketBetaHeaderValue {
 		t.Fatalf("OpenAI-Beta = %s, want %s", got, codexResponsesWebsocketBetaHeaderValue)
 	}
-	if got := headers.Get("User-Agent"); got != codexUserAgent {
-		t.Fatalf("User-Agent = %s, want %s", got, codexUserAgent)
+	if got := headers.Get("User-Agent"); got != "" {
+		t.Fatalf("User-Agent = %s, want empty", got)
 	}
 	if got := headers.Get("Version"); got != "" {
 		t.Fatalf("Version = %q, want empty", got)
@@ -97,8 +97,8 @@ func TestApplyCodexWebsocketHeadersUsesConfigDefaultsForOAuth(t *testing.T) {
 
 	headers := applyCodexWebsocketHeaders(context.Background(), http.Header{}, auth, "", cfg)
 
-	if got := headers.Get("User-Agent"); got != "my-codex-client/1.0" {
-		t.Fatalf("User-Agent = %s, want %s", got, "my-codex-client/1.0")
+	if got := headers.Get("User-Agent"); got != "" {
+		t.Fatalf("User-Agent = %s, want empty", got)
 	}
 	if got := headers.Get("x-codex-beta-features"); got != "feature-a,feature-b" {
 		t.Fatalf("x-codex-beta-features = %s, want %s", got, "feature-a,feature-b")
@@ -129,8 +129,8 @@ func TestApplyCodexWebsocketHeadersPrefersExistingHeadersOverClientAndConfig(t *
 
 	got := applyCodexWebsocketHeaders(ctx, headers, auth, "", cfg)
 
-	if gotVal := got.Get("User-Agent"); gotVal != "existing-ua" {
-		t.Fatalf("User-Agent = %s, want %s", gotVal, "existing-ua")
+	if gotVal := got.Get("User-Agent"); gotVal != "" {
+		t.Fatalf("User-Agent = %s, want empty", gotVal)
 	}
 	if gotVal := got.Get("x-codex-beta-features"); gotVal != "existing-beta" {
 		t.Fatalf("x-codex-beta-features = %s, want %s", gotVal, "existing-beta")
@@ -155,8 +155,8 @@ func TestApplyCodexWebsocketHeadersConfigUserAgentOverridesClientHeader(t *testi
 
 	headers := applyCodexWebsocketHeaders(ctx, http.Header{}, auth, "", cfg)
 
-	if got := headers.Get("User-Agent"); got != "config-ua" {
-		t.Fatalf("User-Agent = %s, want %s", got, "config-ua")
+	if got := headers.Get("User-Agent"); got != "" {
+		t.Fatalf("User-Agent = %s, want empty", got)
 	}
 	if got := headers.Get("x-codex-beta-features"); got != "client-beta" {
 		t.Fatalf("x-codex-beta-features = %s, want %s", got, "client-beta")
@@ -177,8 +177,8 @@ func TestApplyCodexWebsocketHeadersIgnoresConfigForAPIKeyAuth(t *testing.T) {
 
 	headers := applyCodexWebsocketHeaders(context.Background(), http.Header{}, auth, "sk-test", cfg)
 
-	if got := headers.Get("User-Agent"); got != codexUserAgent {
-		t.Fatalf("User-Agent = %s, want %s", got, codexUserAgent)
+	if got := headers.Get("User-Agent"); got != "" {
+		t.Fatalf("User-Agent = %s, want empty", got)
 	}
 	if got := headers.Get("x-codex-beta-features"); got != "" {
 		t.Fatalf("x-codex-beta-features = %q, want empty", got)

internal/runtime/executor/compat_helpers.go

@@ -0,0 +1,129 @@
+package executor
+
+import (
+	"context"
+	"net/http"
+	"time"
+
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
+	"github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/usage"
+	"github.com/tidwall/gjson"
+	"github.com/tiktoken-go/tokenizer"
+)
+
+func newProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
+	return helps.NewProxyAwareHTTPClient(ctx, cfg, auth, timeout)
+}
+
+func parseOpenAIUsage(data []byte) usage.Detail {
+	return helps.ParseOpenAIUsage(data)
+}
+
+func parseOpenAIStreamUsage(line []byte) (usage.Detail, bool) {
+	return helps.ParseOpenAIStreamUsage(line)
+}
+
+func parseOpenAIResponsesUsage(data []byte) usage.Detail {
+	return helps.ParseOpenAIUsage(data)
+}
+
+func parseOpenAIResponsesStreamUsage(line []byte) (usage.Detail, bool) {
+	return helps.ParseOpenAIStreamUsage(line)
+}
+
+func getTokenizer(model string) (tokenizer.Codec, error) {
+	return helps.TokenizerForModel(model)
+}
+
+func countOpenAIChatTokens(enc tokenizer.Codec, payload []byte) (int64, error) {
+	return helps.CountOpenAIChatTokens(enc, payload)
+}
+
+func countClaudeChatTokens(enc tokenizer.Codec, payload []byte) (int64, error) {
+	return helps.CountClaudeChatTokens(enc, payload)
+}
+
+func buildOpenAIUsageJSON(count int64) []byte {
+	return helps.BuildOpenAIUsageJSON(count)
+}
+
+type upstreamRequestLog = helps.UpstreamRequestLog
+
+func recordAPIRequest(ctx context.Context, cfg *config.Config, info upstreamRequestLog) {
+	helps.RecordAPIRequest(ctx, cfg, info)
+}
+
+func recordAPIResponseMetadata(ctx context.Context, cfg *config.Config, status int, headers http.Header) {
+	helps.RecordAPIResponseMetadata(ctx, cfg, status, headers)
+}
+
+func recordAPIResponseError(ctx context.Context, cfg *config.Config, err error) {
+	helps.RecordAPIResponseError(ctx, cfg, err)
+}
+
+func appendAPIResponseChunk(ctx context.Context, cfg *config.Config, chunk []byte) {
+	helps.AppendAPIResponseChunk(ctx, cfg, chunk)
+}
+
+func payloadRequestedModel(opts cliproxyexecutor.Options, fallback string) string {
+	return helps.PayloadRequestedModel(opts, fallback)
+}
+
+func applyPayloadConfigWithRoot(cfg *config.Config, model, protocol, root string, payload, original []byte, requestedModel string) []byte {
+	return helps.ApplyPayloadConfigWithRoot(cfg, model, protocol, root, payload, original, requestedModel)
+}
+
+func summarizeErrorBody(contentType string, body []byte) string {
+	return helps.SummarizeErrorBody(contentType, body)
+}
+
+func apiKeyFromContext(ctx context.Context) string {
+	return helps.APIKeyFromContext(ctx)
+}
+
+func tokenizerForModel(model string) (tokenizer.Codec, error) {
+	return helps.TokenizerForModel(model)
+}
+
+func collectOpenAIContent(content gjson.Result, segments *[]string) {
+	helps.CollectOpenAIContent(content, segments)
+}
+
+type usageReporter struct {
+	reporter *helps.UsageReporter
+}
+
+func newUsageReporter(ctx context.Context, provider, model string, auth *cliproxyauth.Auth) *usageReporter {
+	return &usageReporter{reporter: helps.NewUsageReporter(ctx, provider, model, auth)}
+}
+
+func (r *usageReporter) publish(ctx context.Context, detail usage.Detail) {
+	if r == nil || r.reporter == nil {
+		return
+	}
+	r.reporter.Publish(ctx, detail)
+}
+
+func (r *usageReporter) publishFailure(ctx context.Context) {
+	if r == nil || r.reporter == nil {
+		return
+	}
+	r.reporter.PublishFailure(ctx)
+}
+
+func (r *usageReporter) trackFailure(ctx context.Context, errPtr *error) {
+	if r == nil || r.reporter == nil {
+		return
+	}
+	r.reporter.TrackFailure(ctx, errPtr)
+}
+
+func (r *usageReporter) ensurePublished(ctx context.Context) {
+	if r == nil || r.reporter == nil {
+		return
+	}
+	r.reporter.EnsurePublished(ctx)
+}

internal/runtime/executor/gemini_cli_executor.go

@@ -18,6 +18,7 @@ import (
 
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/misc"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/geminicli"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
@@ -81,6 +82,11 @@ func (e *GeminiCLIExecutor) PrepareRequest(req *http.Request, auth *cliproxyauth
 	}
 	req.Header.Set("Authorization", "Bearer "+tok.AccessToken)
 	applyGeminiCLIHeaders(req, "unknown")
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(req, attrs)
 	return nil
 }
 
@@ -112,8 +118,8 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 		return resp, err
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini-cli")
@@ -132,8 +138,8 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 	}
 
 	basePayload = fixGeminiCLIImageAspectRatio(baseModel, basePayload)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	basePayload = applyPayloadConfigWithRoot(e.cfg, baseModel, "gemini", "request", basePayload, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	basePayload = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, "gemini", "request", basePayload, originalTranslated, requestedModel)
 
 	action := "generateContent"
 	if req.Metadata != nil {
@@ -190,7 +196,8 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 		reqHTTP.Header.Set("Authorization", "Bearer "+tok.AccessToken)
 		applyGeminiCLIHeaders(reqHTTP, attemptModel)
 		reqHTTP.Header.Set("Accept", "application/json")
-		recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+		util.ApplyCustomHeadersFromAttrs(reqHTTP, auth.Attributes)
+		helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 			URL:       url,
 			Method:    http.MethodPost,
 			Headers:   reqHTTP.Header.Clone(),
@@ -204,7 +211,7 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 
 		httpResp, errDo := httpClient.Do(reqHTTP)
 		if errDo != nil {
-			recordAPIResponseError(ctx, e.cfg, errDo)
+			helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 			err = errDo
 			return resp, err
 		}
@@ -213,15 +220,15 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("gemini cli executor: close response body error: %v", errClose)
 		}
-		recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+		helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 		if errRead != nil {
-			recordAPIResponseError(ctx, e.cfg, errRead)
+			helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 			err = errRead
 			return resp, err
 		}
-		appendAPIResponseChunk(ctx, e.cfg, data)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 		if httpResp.StatusCode >= 200 && httpResp.StatusCode < 300 {
-			reporter.publish(ctx, parseGeminiCLIUsage(data))
+			reporter.Publish(ctx, helps.ParseGeminiCLIUsage(data))
 			var param any
 			out := sdktranslator.TranslateNonStream(respCtx, to, from, attemptModel, opts.OriginalRequest, payload, data, &param)
 			resp = cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}
@@ -230,7 +237,7 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 
 		lastStatus = httpResp.StatusCode
 		lastBody = append([]byte(nil), data...)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
 		if httpResp.StatusCode == 429 {
 			if idx+1 < len(models) {
 				log.Debugf("gemini cli executor: rate limited, retrying with next model: %s", models[idx+1])
@@ -245,7 +252,7 @@ func (e *GeminiCLIExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth
 	}
 
 	if len(lastBody) > 0 {
-		appendAPIResponseChunk(ctx, e.cfg, lastBody)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, lastBody)
 	}
 	if lastStatus == 0 {
 		lastStatus = 429
@@ -266,8 +273,8 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 		return nil, err
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini-cli")
@@ -286,8 +293,8 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 	}
 
 	basePayload = fixGeminiCLIImageAspectRatio(baseModel, basePayload)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	basePayload = applyPayloadConfigWithRoot(e.cfg, baseModel, "gemini", "request", basePayload, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	basePayload = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, "gemini", "request", basePayload, originalTranslated, requestedModel)
 
 	projectID := resolveGeminiProjectID(auth)
 
@@ -335,7 +342,8 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 		reqHTTP.Header.Set("Authorization", "Bearer "+tok.AccessToken)
 		applyGeminiCLIHeaders(reqHTTP, attemptModel)
 		reqHTTP.Header.Set("Accept", "text/event-stream")
-		recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+		util.ApplyCustomHeadersFromAttrs(reqHTTP, auth.Attributes)
+		helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 			URL:       url,
 			Method:    http.MethodPost,
 			Headers:   reqHTTP.Header.Clone(),
@@ -349,25 +357,25 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 
 		httpResp, errDo := httpClient.Do(reqHTTP)
 		if errDo != nil {
-			recordAPIResponseError(ctx, e.cfg, errDo)
+			helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 			err = errDo
 			return nil, err
 		}
-		recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+		helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 		if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 			data, errRead := io.ReadAll(httpResp.Body)
 			if errClose := httpResp.Body.Close(); errClose != nil {
 				log.Errorf("gemini cli executor: close response body error: %v", errClose)
 			}
 			if errRead != nil {
-				recordAPIResponseError(ctx, e.cfg, errRead)
+				helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 				err = errRead
 				return nil, err
 			}
-			appendAPIResponseChunk(ctx, e.cfg, data)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 			lastStatus = httpResp.StatusCode
 			lastBody = append([]byte(nil), data...)
-			logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
+			helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
 			if httpResp.StatusCode == 429 {
 				if idx+1 < len(models) {
 					log.Debugf("gemini cli executor: rate limited, retrying with next model: %s", models[idx+1])
@@ -394,9 +402,9 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 				var param any
 				for scanner.Scan() {
 					line := scanner.Bytes()
-					appendAPIResponseChunk(ctx, e.cfg, line)
-					if detail, ok := parseGeminiCLIStreamUsage(line); ok {
-						reporter.publish(ctx, detail)
+					helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+					if detail, ok := helps.ParseGeminiCLIStreamUsage(line); ok {
+						reporter.Publish(ctx, detail)
 					}
 					if bytes.HasPrefix(line, dataTag) {
 						segments := sdktranslator.TranslateStream(respCtx, to, from, attemptModel, opts.OriginalRequest, reqBody, bytes.Clone(line), &param)
@@ -411,8 +419,8 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 					out <- cliproxyexecutor.StreamChunk{Payload: segments[i]}
 				}
 				if errScan := scanner.Err(); errScan != nil {
-					recordAPIResponseError(ctx, e.cfg, errScan)
-					reporter.publishFailure(ctx)
+					helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+					reporter.PublishFailure(ctx)
 					out <- cliproxyexecutor.StreamChunk{Err: errScan}
 				}
 				return
@@ -420,13 +428,13 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 
 			data, errRead := io.ReadAll(resp.Body)
 			if errRead != nil {
-				recordAPIResponseError(ctx, e.cfg, errRead)
-				reporter.publishFailure(ctx)
+				helps.RecordAPIResponseError(ctx, e.cfg, errRead)
+				reporter.PublishFailure(ctx)
 				out <- cliproxyexecutor.StreamChunk{Err: errRead}
 				return
 			}
-			appendAPIResponseChunk(ctx, e.cfg, data)
-			reporter.publish(ctx, parseGeminiCLIUsage(data))
+			helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+			reporter.Publish(ctx, helps.ParseGeminiCLIUsage(data))
 			var param any
 			segments := sdktranslator.TranslateStream(respCtx, to, from, attemptModel, opts.OriginalRequest, reqBody, data, &param)
 			for i := range segments {
@@ -443,7 +451,7 @@ func (e *GeminiCLIExecutor) ExecuteStream(ctx context.Context, auth *cliproxyaut
 	}
 
 	if len(lastBody) > 0 {
-		appendAPIResponseChunk(ctx, e.cfg, lastBody)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, lastBody)
 	}
 	if lastStatus == 0 {
 		lastStatus = 429
@@ -516,7 +524,8 @@ func (e *GeminiCLIExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.
 		reqHTTP.Header.Set("Authorization", "Bearer "+tok.AccessToken)
 		applyGeminiCLIHeaders(reqHTTP, baseModel)
 		reqHTTP.Header.Set("Accept", "application/json")
-		recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+		util.ApplyCustomHeadersFromAttrs(reqHTTP, auth.Attributes)
+		helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 			URL:       url,
 			Method:    http.MethodPost,
 			Headers:   reqHTTP.Header.Clone(),
@@ -530,17 +539,19 @@ func (e *GeminiCLIExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.
 
 		resp, errDo := httpClient.Do(reqHTTP)
 		if errDo != nil {
-			recordAPIResponseError(ctx, e.cfg, errDo)
+			helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 			return cliproxyexecutor.Response{}, errDo
 		}
 		data, errRead := io.ReadAll(resp.Body)
-		_ = resp.Body.Close()
-		recordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
+		if errClose := resp.Body.Close(); errClose != nil {
+			helps.LogWithRequestID(ctx).Errorf("response body close error: %v", errClose)
+		}
+		helps.RecordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
 		if errRead != nil {
-			recordAPIResponseError(ctx, e.cfg, errRead)
+			helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 			return cliproxyexecutor.Response{}, errRead
 		}
-		appendAPIResponseChunk(ctx, e.cfg, data)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 		if resp.StatusCode >= 200 && resp.StatusCode < 300 {
 			count := gjson.GetBytes(data, "totalTokens").Int()
 			translated := sdktranslator.TranslateTokenCount(respCtx, to, from, count, data)
@@ -611,7 +622,7 @@ func prepareGeminiCLITokenSource(ctx context.Context, cfg *config.Config, auth *
 	}
 
 	ctxToken := ctx
-	if httpClient := newProxyAwareHTTPClient(ctx, cfg, auth, 0); httpClient != nil {
+	if httpClient := helps.NewProxyAwareHTTPClient(ctx, cfg, auth, 0); httpClient != nil {
 		ctxToken = context.WithValue(ctxToken, oauth2.HTTPClient, httpClient)
 	}
 
@@ -707,7 +718,7 @@ func geminiOAuthMetadata(auth *cliproxyauth.Auth) map[string]any {
 }
 
 func newHTTPClient(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
-	return newProxyAwareHTTPClient(ctx, cfg, auth, timeout)
+	return helps.NewProxyAwareHTTPClient(ctx, cfg, auth, timeout)
 }
 
 func cloneMap(in map[string]any) map[string]any {

internal/runtime/executor/gemini_executor.go

@@ -13,6 +13,7 @@ import (
 	"strings"
 
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -85,7 +86,7 @@ func (e *GeminiExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Aut
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -110,8 +111,8 @@ func (e *GeminiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 
 	apiKey, bearer := geminiCreds(auth)
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	// Official Gemini API via API key or OAuth bearer
 	from := opts.SourceFormat
@@ -130,8 +131,8 @@ func (e *GeminiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 	}
 
 	body = fixGeminiImageAspectRatio(baseModel, body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 
 	action := "generateContent"
@@ -165,7 +166,7 @@ func (e *GeminiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -177,10 +178,10 @@ func (e *GeminiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -188,21 +189,21 @@ func (e *GeminiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, r
 			log.Errorf("gemini executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseGeminiUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseGeminiUsage(data))
 	var param any
 	out := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, body, data, &param)
 	resp = cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}
@@ -218,8 +219,8 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 
 	apiKey, bearer := geminiCreds(auth)
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini")
@@ -237,8 +238,8 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 	}
 
 	body = fixGeminiImageAspectRatio(baseModel, body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 
 	baseURL := resolveGeminiBaseURL(auth)
@@ -268,7 +269,7 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -280,17 +281,17 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("gemini executor: close response body error: %v", errClose)
 		}
@@ -310,14 +311,14 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			filtered := FilterSSEUsageMetadata(line)
-			payload := jsonPayload(filtered)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			filtered := helps.FilterSSEUsageMetadata(line)
+			payload := helps.JSONPayload(filtered)
 			if len(payload) == 0 {
 				continue
 			}
-			if detail, ok := parseGeminiStreamUsage(payload); ok {
-				reporter.publish(ctx, detail)
+			if detail, ok := helps.ParseGeminiStreamUsage(payload); ok {
+				reporter.Publish(ctx, detail)
 			}
 			lines := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(payload), &param)
 			for i := range lines {
@@ -329,8 +330,8 @@ func (e *GeminiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.A
 			out <- cliproxyexecutor.StreamChunk{Payload: lines[i]}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -381,7 +382,7 @@ func (e *GeminiExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -393,23 +394,27 @@ func (e *GeminiExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Aut
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	resp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return cliproxyexecutor.Response{}, err
 	}
-	defer func() { _ = resp.Body.Close() }()
-	recordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
+	defer func() {
+		if errClose := resp.Body.Close(); errClose != nil {
+			helps.LogWithRequestID(ctx).Errorf("response body close error: %v", errClose)
+		}
+	}()
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, resp.StatusCode, resp.Header.Clone())
 
 	data, err := io.ReadAll(resp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return cliproxyexecutor.Response{}, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", resp.StatusCode, summarizeErrorBody(resp.Header.Get("Content-Type"), data))
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", resp.StatusCode, helps.SummarizeErrorBody(resp.Header.Get("Content-Type"), data))
 		return cliproxyexecutor.Response{}, statusErr{code: resp.StatusCode, msg: string(data)}
 	}
 

internal/runtime/executor/gemini_vertex_executor.go

@@ -16,7 +16,9 @@ import (
 
 	vertexauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/vertex"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
@@ -227,7 +229,7 @@ func (e *GeminiVertexExecutor) HttpRequest(ctx context.Context, auth *cliproxyau
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -301,8 +303,8 @@ func (e *GeminiVertexExecutor) Refresh(_ context.Context, auth *cliproxyauth.Aut
 func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options, projectID, location string, saJSON []byte) (resp cliproxyexecutor.Response, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	var body []byte
 
@@ -332,8 +334,8 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 		}
 
 		body = fixGeminiImageAspectRatio(baseModel, body)
-		requestedModel := payloadRequestedModel(opts, req.Model)
-		body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+		requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+		body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 		body, _ = sjson.SetBytes(body, "model", baseModel)
 	}
 
@@ -362,6 +364,11 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 		return resp, statusErr{code: 500, msg: "internal server error"}
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -369,7 +376,7 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -381,10 +388,10 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return resp, errDo
 	}
 	defer func() {
@@ -392,21 +399,21 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 	data, errRead := io.ReadAll(httpResp.Body)
 	if errRead != nil {
-		recordAPIResponseError(ctx, e.cfg, errRead)
+		helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 		return resp, errRead
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseGeminiUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseGeminiUsage(data))
 
 	// For Imagen models, convert response to Gemini format before translation
 	// This ensures Imagen responses use the same format as gemini-3-pro-image-preview
@@ -427,8 +434,8 @@ func (e *GeminiVertexExecutor) executeWithServiceAccount(ctx context.Context, au
 func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options, apiKey, baseURL string) (resp cliproxyexecutor.Response, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini")
@@ -447,8 +454,8 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 	}
 
 	body = fixGeminiImageAspectRatio(baseModel, body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 
 	action := getVertexAction(baseModel, false)
@@ -477,6 +484,11 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 		httpReq.Header.Set("x-goog-api-key", apiKey)
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -484,7 +496,7 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -496,10 +508,10 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return resp, errDo
 	}
 	defer func() {
@@ -507,21 +519,21 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 	data, errRead := io.ReadAll(httpResp.Body)
 	if errRead != nil {
-		recordAPIResponseError(ctx, e.cfg, errRead)
+		helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 		return resp, errRead
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseGeminiUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseGeminiUsage(data))
 	var param any
 	out := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, body, data, &param)
 	resp = cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}
@@ -532,8 +544,8 @@ func (e *GeminiVertexExecutor) executeWithAPIKey(ctx context.Context, auth *clip
 func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options, projectID, location string, saJSON []byte) (_ *cliproxyexecutor.StreamResult, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini")
@@ -552,8 +564,8 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 	}
 
 	body = fixGeminiImageAspectRatio(baseModel, body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 
 	action := getVertexAction(baseModel, true)
@@ -581,6 +593,11 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 		return nil, statusErr{code: 500, msg: "internal server error"}
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -588,7 +605,7 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -600,17 +617,17 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return nil, errDo
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
@@ -630,9 +647,9 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseGeminiStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseGeminiStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			lines := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(line), &param)
 			for i := range lines {
@@ -644,8 +661,8 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 			out <- cliproxyexecutor.StreamChunk{Payload: lines[i]}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -656,8 +673,8 @@ func (e *GeminiVertexExecutor) executeStreamWithServiceAccount(ctx context.Conte
 func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options, apiKey, baseURL string) (_ *cliproxyexecutor.StreamResult, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("gemini")
@@ -676,8 +693,8 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 	}
 
 	body = fixGeminiImageAspectRatio(baseModel, body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, _ = sjson.SetBytes(body, "model", baseModel)
 
 	action := getVertexAction(baseModel, true)
@@ -705,6 +722,11 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 		httpReq.Header.Set("x-goog-api-key", apiKey)
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -712,7 +734,7 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -724,17 +746,17 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return nil, errDo
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
@@ -754,9 +776,9 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseGeminiStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseGeminiStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			lines := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(line), &param)
 			for i := range lines {
@@ -768,8 +790,8 @@ func (e *GeminiVertexExecutor) executeStreamWithAPIKey(ctx context.Context, auth
 			out <- cliproxyexecutor.StreamChunk{Payload: lines[i]}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -812,6 +834,11 @@ func (e *GeminiVertexExecutor) countTokensWithServiceAccount(ctx context.Context
 		return cliproxyexecutor.Response{}, statusErr{code: 500, msg: "internal server error"}
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -819,7 +846,7 @@ func (e *GeminiVertexExecutor) countTokensWithServiceAccount(ctx context.Context
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -831,10 +858,10 @@ func (e *GeminiVertexExecutor) countTokensWithServiceAccount(ctx context.Context
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return cliproxyexecutor.Response{}, errDo
 	}
 	defer func() {
@@ -842,19 +869,19 @@ func (e *GeminiVertexExecutor) countTokensWithServiceAccount(ctx context.Context
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		return cliproxyexecutor.Response{}, statusErr{code: httpResp.StatusCode, msg: string(b)}
 	}
 	data, errRead := io.ReadAll(httpResp.Body)
 	if errRead != nil {
-		recordAPIResponseError(ctx, e.cfg, errRead)
+		helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 		return cliproxyexecutor.Response{}, errRead
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 	count := gjson.GetBytes(data, "totalTokens").Int()
 	out := sdktranslator.TranslateTokenCount(ctx, to, from, count, data)
 	return cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}, nil
@@ -896,6 +923,11 @@ func (e *GeminiVertexExecutor) countTokensWithAPIKey(ctx context.Context, auth *
 		httpReq.Header.Set("x-goog-api-key", apiKey)
 	}
 	applyGeminiHeaders(httpReq, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
@@ -903,7 +935,7 @@ func (e *GeminiVertexExecutor) countTokensWithAPIKey(ctx context.Context, auth *
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -915,10 +947,10 @@ func (e *GeminiVertexExecutor) countTokensWithAPIKey(ctx context.Context, auth *
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, errDo := httpClient.Do(httpReq)
 	if errDo != nil {
-		recordAPIResponseError(ctx, e.cfg, errDo)
+		helps.RecordAPIResponseError(ctx, e.cfg, errDo)
 		return cliproxyexecutor.Response{}, errDo
 	}
 	defer func() {
@@ -926,19 +958,19 @@ func (e *GeminiVertexExecutor) countTokensWithAPIKey(ctx context.Context, auth *
 			log.Errorf("vertex executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		return cliproxyexecutor.Response{}, statusErr{code: httpResp.StatusCode, msg: string(b)}
 	}
 	data, errRead := io.ReadAll(httpResp.Body)
 	if errRead != nil {
-		recordAPIResponseError(ctx, e.cfg, errRead)
+		helps.RecordAPIResponseError(ctx, e.cfg, errRead)
 		return cliproxyexecutor.Response{}, errRead
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
 	count := gjson.GetBytes(data, "totalTokens").Int()
 	out := sdktranslator.TranslateTokenCount(ctx, to, from, count, data)
 	return cliproxyexecutor.Response{Payload: out, Headers: httpResp.Header.Clone()}, nil
@@ -1012,7 +1044,7 @@ func vertexBaseURL(location string) string {
 }
 
 func vertexAccessToken(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, saJSON []byte) (string, error) {
-	if httpClient := newProxyAwareHTTPClient(ctx, cfg, auth, 0); httpClient != nil {
+	if httpClient := helps.NewProxyAwareHTTPClient(ctx, cfg, auth, 0); httpClient != nil {
 		ctx = context.WithValue(ctx, oauth2.HTTPClient, httpClient)
 	}
 	// Use cloud-platform scope for Vertex AI.

internal/runtime/executor/github_copilot_executor.go

@@ -7,6 +7,7 @@ import (
 	"fmt"
 	"io"
 	"net/http"
+	"slices"
 	"strings"
 	"sync"
 	"time"
@@ -17,6 +18,7 @@ import (
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
 	log "github.com/sirupsen/logrus"
@@ -40,7 +42,7 @@ const (
 	copilotEditorVersion = "vscode/1.107.0"
 	copilotPluginVersion = "copilot-chat/0.35.0"
 	copilotIntegrationID = "vscode-chat"
-	copilotOpenAIIntent  = "conversation-panel"
+	copilotOpenAIIntent  = "conversation-edits"
 	copilotGitHubAPIVer  = "2025-04-01"
 )
 
@@ -126,6 +128,7 @@ func (e *GitHubCopilotExecutor) Execute(ctx context.Context, auth *cliproxyauth.
 	body := sdktranslator.TranslateRequest(from, to, req.Model, bytes.Clone(req.Payload), false)
 	body = e.normalizeModel(req.Model, body)
 	body = flattenAssistantContent(body)
+	body = stripUnsupportedBetas(body)
 
 	// Detect vision content before input normalization removes messages
 	hasVision := detectVisionContent(body)
@@ -142,6 +145,7 @@ func (e *GitHubCopilotExecutor) Execute(ctx context.Context, auth *cliproxyauth.
 	if useResponses {
 		body = normalizeGitHubCopilotResponsesInput(body)
 		body = normalizeGitHubCopilotResponsesTools(body)
+		body = applyGitHubCopilotResponsesDefaults(body)
 	} else {
 		body = normalizeGitHubCopilotChatTools(body)
 	}
@@ -225,9 +229,10 @@ func (e *GitHubCopilotExecutor) Execute(ctx context.Context, auth *cliproxyauth.
 	if useResponses && from.String() == "claude" {
 		converted = translateGitHubCopilotResponsesNonStreamToClaude(data)
 	} else {
+		data = normalizeGitHubCopilotReasoningField(data)
 		converted = sdktranslator.TranslateNonStream(ctx, to, from, req.Model, bytes.Clone(opts.OriginalRequest), body, data, &param)
 	}
-	resp = cliproxyexecutor.Response{Payload: converted}
+	resp = cliproxyexecutor.Response{Payload: converted, Headers: httpResp.Header.Clone()}
 	reporter.ensurePublished(ctx)
 	return resp, nil
 }
@@ -256,6 +261,7 @@ func (e *GitHubCopilotExecutor) ExecuteStream(ctx context.Context, auth *cliprox
 	body := sdktranslator.TranslateRequest(from, to, req.Model, bytes.Clone(req.Payload), true)
 	body = e.normalizeModel(req.Model, body)
 	body = flattenAssistantContent(body)
+	body = stripUnsupportedBetas(body)
 
 	// Detect vision content before input normalization removes messages
 	hasVision := detectVisionContent(body)
@@ -272,6 +278,7 @@ func (e *GitHubCopilotExecutor) ExecuteStream(ctx context.Context, auth *cliprox
 	if useResponses {
 		body = normalizeGitHubCopilotResponsesInput(body)
 		body = normalizeGitHubCopilotResponsesTools(body)
+		body = applyGitHubCopilotResponsesDefaults(body)
 	} else {
 		body = normalizeGitHubCopilotChatTools(body)
 	}
@@ -378,7 +385,20 @@ func (e *GitHubCopilotExecutor) ExecuteStream(ctx context.Context, auth *cliprox
 			if useResponses && from.String() == "claude" {
 				chunks = translateGitHubCopilotResponsesStreamToClaude(bytes.Clone(line), &param)
 			} else {
-				chunks = sdktranslator.TranslateStream(ctx, to, from, req.Model, bytes.Clone(opts.OriginalRequest), body, bytes.Clone(line), &param)
+				// Strip SSE "data: " prefix before reasoning field normalization,
+				// since normalizeGitHubCopilotReasoningField expects pure JSON.
+				// Re-wrap with the prefix afterward for the translator.
+				normalizedLine := bytes.Clone(line)
+				if bytes.HasPrefix(line, dataTag) {
+					sseData := bytes.TrimSpace(line[len(dataTag):])
+					if !bytes.Equal(sseData, []byte("[DONE]")) && gjson.ValidBytes(sseData) {
+						normalized := normalizeGitHubCopilotReasoningField(bytes.Clone(sseData))
+						if !bytes.Equal(normalized, sseData) {
+							normalizedLine = append(append([]byte(nil), dataTag...), normalized...)
+						}
+					}
+				}
+				chunks = sdktranslator.TranslateStream(ctx, to, from, req.Model, bytes.Clone(opts.OriginalRequest), body, normalizedLine, &param)
 			}
 			for i := range chunks {
 				out <- cliproxyexecutor.StreamChunk{Payload: bytes.Clone(chunks[i])}
@@ -400,9 +420,28 @@ func (e *GitHubCopilotExecutor) ExecuteStream(ctx context.Context, auth *cliprox
 	}, nil
 }
 
-// CountTokens is not supported for GitHub Copilot.
-func (e *GitHubCopilotExecutor) CountTokens(_ context.Context, _ *cliproxyauth.Auth, _ cliproxyexecutor.Request, _ cliproxyexecutor.Options) (cliproxyexecutor.Response, error) {
-	return cliproxyexecutor.Response{}, statusErr{code: http.StatusNotImplemented, msg: "count tokens not supported for github-copilot"}
+// CountTokens estimates token count locally using tiktoken, since the GitHub
+// Copilot API does not expose a dedicated token counting endpoint.
+func (e *GitHubCopilotExecutor) CountTokens(ctx context.Context, _ *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options) (cliproxyexecutor.Response, error) {
+	baseModel := thinking.ParseSuffix(req.Model).ModelName
+
+	from := opts.SourceFormat
+	to := sdktranslator.FromString("openai")
+	translated := sdktranslator.TranslateRequest(from, to, baseModel, req.Payload, false)
+
+	enc, err := helps.TokenizerForModel(baseModel)
+	if err != nil {
+		return cliproxyexecutor.Response{}, fmt.Errorf("github copilot executor: tokenizer init failed: %w", err)
+	}
+
+	count, err := helps.CountOpenAIChatTokens(enc, translated)
+	if err != nil {
+		return cliproxyexecutor.Response{}, fmt.Errorf("github copilot executor: token counting failed: %w", err)
+	}
+
+	usageJSON := helps.BuildOpenAIUsageJSON(count)
+	translatedUsage := sdktranslator.TranslateTokenCount(ctx, to, from, count, usageJSON)
+	return cliproxyexecutor.Response{Payload: translatedUsage}, nil
 }
 
 // Refresh validates the GitHub token is still working.
@@ -491,46 +530,127 @@ func (e *GitHubCopilotExecutor) applyHeaders(r *http.Request, apiToken string, b
 	r.Header.Set("X-Request-Id", uuid.NewString())
 
 	initiator := "user"
-	if role := detectLastConversationRole(body); role == "assistant" || role == "tool" {
+	if isAgentInitiated(body) {
 		initiator = "agent"
 	}
 	r.Header.Set("X-Initiator", initiator)
 }
 
-func detectLastConversationRole(body []byte) string {
+// isAgentInitiated determines whether the current request is agent-initiated
+// (tool callbacks, continuations) rather than user-initiated (new user prompt).
+//
+// GitHub Copilot uses the X-Initiator header for billing:
+//   - "user"  → consumes premium request quota
+//   - "agent" → free (tool loops, continuations)
+//
+// The challenge: Claude Code sends tool results as role:"user" messages with
+// content type "tool_result". After translation to OpenAI format, the tool_result
+// part becomes a separate role:"tool" message, but if the original Claude message
+// also contained text content (e.g. skill invocations, attachment descriptions),
+// a role:"user" message is emitted AFTER the tool message, making the last message
+// appear user-initiated when it's actually part of an agent tool loop.
+//
+// VSCode Copilot Chat solves this with explicit flags (iterationNumber,
+// isContinuation, subAgentInvocationId). Since CPA doesn't have these flags,
+// we infer agent status by checking whether the conversation contains prior
+// assistant/tool messages — if it does, the current request is a continuation.
+//
+// References:
+//   - opencode#8030, opencode#15824: same root cause and fix approach
+//   - vscode-copilot-chat: toolCallingLoop.ts (iterationNumber === 0)
+//   - pi-ai: github-copilot-headers.ts (last message role check)
+func isAgentInitiated(body []byte) bool {
 	if len(body) == 0 {
-		return ""
+		return false
 	}
 
+	// Chat Completions API: check messages array
 	if messages := gjson.GetBytes(body, "messages"); messages.Exists() && messages.IsArray() {
 		arr := messages.Array()
+		if len(arr) == 0 {
+			return false
+		}
+
+		lastRole := ""
 		for i := len(arr) - 1; i >= 0; i-- {
-			if role := arr[i].Get("role").String(); role != "" {
-				return role
+			if r := arr[i].Get("role").String(); r != "" {
+				lastRole = r
+				break
 			}
 		}
+
+		// If last message is assistant or tool, clearly agent-initiated.
+		if lastRole == "assistant" || lastRole == "tool" {
+			return true
+		}
+
+		// If last message is "user", check whether it contains tool results
+		// (indicating a tool-loop continuation) or if the preceding message
+		// is an assistant tool_use. This is more precise than checking for
+		// any prior assistant message, which would false-positive on genuine
+		// multi-turn follow-ups.
+		if lastRole == "user" {
+			// Check if the last user message contains tool_result content
+			lastContent := arr[len(arr)-1].Get("content")
+			if lastContent.Exists() && lastContent.IsArray() {
+				for _, part := range lastContent.Array() {
+					if part.Get("type").String() == "tool_result" {
+						return true
+					}
+				}
+			}
+			// Check if the second-to-last message is an assistant with tool_use
+			if len(arr) >= 2 {
+				prev := arr[len(arr)-2]
+				if prev.Get("role").String() == "assistant" {
+					prevContent := prev.Get("content")
+					if prevContent.Exists() && prevContent.IsArray() {
+						for _, part := range prevContent.Array() {
+							if part.Get("type").String() == "tool_use" {
+								return true
+							}
+						}
+					}
+				}
+			}
+		}
+
+		return false
 	}
 
+	// Responses API: check input array
 	if inputs := gjson.GetBytes(body, "input"); inputs.Exists() && inputs.IsArray() {
 		arr := inputs.Array()
-		for i := len(arr) - 1; i >= 0; i-- {
-			item := arr[i]
+		if len(arr) == 0 {
+			return false
+		}
 
-			// Most Responses input items carry a top-level role.
-			if role := item.Get("role").String(); role != "" {
-				return role
+		// Check last item
+		last := arr[len(arr)-1]
+		if role := last.Get("role").String(); role == "assistant" {
+			return true
+		}
+		switch last.Get("type").String() {
+		case "function_call", "function_call_arguments", "computer_call":
+			return true
+		case "function_call_output", "function_call_response", "tool_result", "computer_call_output":
+			return true
+		}
+
+		// If last item is user-role, check for prior non-user items
+		for _, item := range arr {
+			if role := item.Get("role").String(); role == "assistant" {
+				return true
 			}
-
 			switch item.Get("type").String() {
-			case "function_call", "function_call_arguments", "computer_call":
-				return "assistant"
-			case "function_call_output", "function_call_response", "tool_result", "computer_call_output":
-				return "tool"
+			case "function_call", "function_call_output", "function_call_response",
+				"function_call_arguments", "computer_call", "computer_call_output":
+				return true
 			}
 		}
 	}
 
-	return ""
+	return false
 }
 
 // detectVisionContent checks if the request body contains vision/image content.
@@ -572,6 +692,85 @@ func (e *GitHubCopilotExecutor) normalizeModel(model string, body []byte) []byte
 	return body
 }
 
+// copilotUnsupportedBetas lists beta headers that are Anthropic-specific and
+// must not be forwarded to GitHub Copilot. The context-1m beta enables 1M
+// context on Anthropic's API, but Copilot's Claude models are limited to
+// ~128K-200K. Passing it through would not enable 1M on Copilot, but stripping
+// it from the translated body avoids confusing downstream translators.
+var copilotUnsupportedBetas = []string{
+	"context-1m-2025-08-07",
+}
+
+// stripUnsupportedBetas removes Anthropic-specific beta entries from the
+// translated request body. In OpenAI format the betas may appear under
+// "metadata.betas" or a top-level "betas" array; in Claude format they sit at
+// "betas". This function checks all known locations.
+func stripUnsupportedBetas(body []byte) []byte {
+	betaPaths := []string{"betas", "metadata.betas"}
+	for _, path := range betaPaths {
+		arr := gjson.GetBytes(body, path)
+		if !arr.Exists() || !arr.IsArray() {
+			continue
+		}
+		var filtered []string
+		changed := false
+		for _, item := range arr.Array() {
+			beta := item.String()
+			if isCopilotUnsupportedBeta(beta) {
+				changed = true
+				continue
+			}
+			filtered = append(filtered, beta)
+		}
+		if !changed {
+			continue
+		}
+		if len(filtered) == 0 {
+			body, _ = sjson.DeleteBytes(body, path)
+		} else {
+			body, _ = sjson.SetBytes(body, path, filtered)
+		}
+	}
+	return body
+}
+
+func isCopilotUnsupportedBeta(beta string) bool {
+	return slices.Contains(copilotUnsupportedBetas, beta)
+}
+
+// normalizeGitHubCopilotReasoningField maps Copilot's non-standard
+// 'reasoning_text' field to the standard OpenAI 'reasoning_content' field
+// that the SDK translator expects. This handles both streaming deltas
+// (choices[].delta.reasoning_text) and non-streaming messages
+// (choices[].message.reasoning_text). The field is only renamed when
+// 'reasoning_content' is absent or null, preserving standard responses.
+// All choices are processed to support n>1 requests.
+func normalizeGitHubCopilotReasoningField(data []byte) []byte {
+	choices := gjson.GetBytes(data, "choices")
+	if !choices.Exists() || !choices.IsArray() {
+		return data
+	}
+	for i := range choices.Array() {
+		// Non-streaming: choices[i].message.reasoning_text
+		msgRT := fmt.Sprintf("choices.%d.message.reasoning_text", i)
+		msgRC := fmt.Sprintf("choices.%d.message.reasoning_content", i)
+		if rt := gjson.GetBytes(data, msgRT); rt.Exists() && rt.String() != "" {
+			if rc := gjson.GetBytes(data, msgRC); !rc.Exists() || rc.Type == gjson.Null || rc.String() == "" {
+				data, _ = sjson.SetBytes(data, msgRC, rt.String())
+			}
+		}
+		// Streaming: choices[i].delta.reasoning_text
+		deltaRT := fmt.Sprintf("choices.%d.delta.reasoning_text", i)
+		deltaRC := fmt.Sprintf("choices.%d.delta.reasoning_content", i)
+		if rt := gjson.GetBytes(data, deltaRT); rt.Exists() && rt.String() != "" {
+			if rc := gjson.GetBytes(data, deltaRC); !rc.Exists() || rc.Type == gjson.Null || rc.String() == "" {
+				data, _ = sjson.SetBytes(data, deltaRC, rt.String())
+			}
+		}
+	}
+	return data
+}
+
 func useGitHubCopilotResponsesEndpoint(sourceFormat sdktranslator.Format, model string) bool {
 	if sourceFormat.String() == "openai-response" {
 		return true
@@ -596,12 +795,7 @@ func lookupGitHubCopilotStaticModelInfo(model string) *registry.ModelInfo {
 }
 
 func containsEndpoint(endpoints []string, endpoint string) bool {
-	for _, item := range endpoints {
-		if item == endpoint {
-			return true
-		}
-	}
-	return false
+	return slices.Contains(endpoints, endpoint)
 }
 
 // flattenAssistantContent converts assistant message content from array format
@@ -856,6 +1050,32 @@ func stripGitHubCopilotResponsesUnsupportedFields(body []byte) []byte {
 	return body
 }
 
+// applyGitHubCopilotResponsesDefaults sets required fields for the Responses API
+// that both vscode-copilot-chat and pi-ai always include.
+//
+// References:
+//   - vscode-copilot-chat: src/platform/endpoint/node/responsesApi.ts
+//   - pi-ai (badlogic/pi-mono): packages/ai/src/providers/openai-responses.ts
+func applyGitHubCopilotResponsesDefaults(body []byte) []byte {
+	// store: false — prevents request/response storage
+	if !gjson.GetBytes(body, "store").Exists() {
+		body, _ = sjson.SetBytes(body, "store", false)
+	}
+
+	// include: ["reasoning.encrypted_content"] — enables reasoning content
+	// reuse across turns, avoiding redundant computation
+	if !gjson.GetBytes(body, "include").Exists() {
+		body, _ = sjson.SetRawBytes(body, "include", []byte(`["reasoning.encrypted_content"]`))
+	}
+
+	// If reasoning.effort is set but reasoning.summary is not, default to "auto"
+	if gjson.GetBytes(body, "reasoning.effort").Exists() && !gjson.GetBytes(body, "reasoning.summary").Exists() {
+		body, _ = sjson.SetBytes(body, "reasoning.summary", "auto")
+	}
+
+	return body
+}
+
 func normalizeGitHubCopilotResponsesTools(body []byte) []byte {
 	tools := gjson.GetBytes(body, "tools")
 	if tools.Exists() {
@@ -1406,6 +1626,21 @@ func FetchGitHubCopilotModels(ctx context.Context, auth *cliproxyauth.Auth, cfg
 			m.MaxCompletionTokens = defaultCopilotMaxCompletionTokens
 		}
 
+		// Override with real limits from the Copilot API when available.
+		// The API returns per-account limits (individual vs business) under
+		// capabilities.limits, which are more accurate than our static
+		// fallback values. We use max_prompt_tokens as ContextLength because
+		// that's the hard limit the Copilot API enforces on prompt size —
+		// exceeding it triggers "prompt token count exceeds the limit" errors.
+		if limits := entry.Limits(); limits != nil {
+			if limits.MaxPromptTokens > 0 {
+				m.ContextLength = limits.MaxPromptTokens
+			}
+			if limits.MaxOutputTokens > 0 {
+				m.MaxCompletionTokens = limits.MaxOutputTokens
+			}
+		}
+
 		models = append(models, m)
 	}
 

internal/runtime/executor/github_copilot_executor_test.go

@@ -1,11 +1,14 @@
 package executor
 
 import (
+	"context"
 	"net/http"
 	"strings"
 	"testing"
 
+	copilotauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/copilot"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
 	"github.com/tidwall/gjson"
 )
@@ -72,26 +75,39 @@ func TestUseGitHubCopilotResponsesEndpoint_CodexModel(t *testing.T) {
 }
 
 func TestUseGitHubCopilotResponsesEndpoint_RegistryResponsesOnlyModel(t *testing.T) {
-	t.Parallel()
+	// Not parallel: shares global model registry with DynamicRegistryWinsOverStatic.
 	if !useGitHubCopilotResponsesEndpoint(sdktranslator.FromString("openai"), "gpt-5.4") {
 		t.Fatal("expected responses-only registry model to use /responses")
 	}
+	if !useGitHubCopilotResponsesEndpoint(sdktranslator.FromString("openai"), "gpt-5.4-mini") {
+		t.Fatal("expected responses-only registry model to use /responses")
+	}
 }
 
 func TestUseGitHubCopilotResponsesEndpoint_DynamicRegistryWinsOverStatic(t *testing.T) {
-	t.Parallel()
+	// Not parallel: mutates global model registry, conflicts with RegistryResponsesOnlyModel.
 
 	reg := registry.GetGlobalRegistry()
 	clientID := "github-copilot-test-client"
-	reg.RegisterClient(clientID, "github-copilot", []*registry.ModelInfo{{
-		ID:                 "gpt-5.4",
-		SupportedEndpoints: []string{"/chat/completions", "/responses"},
-	}})
+	reg.RegisterClient(clientID, "github-copilot", []*registry.ModelInfo{
+		{
+			ID:                 "gpt-5.4",
+			SupportedEndpoints: []string{"/chat/completions", "/responses"},
+		},
+		{
+			ID:                 "gpt-5.4-mini",
+			SupportedEndpoints: []string{"/chat/completions", "/responses"},
+		},
+	})
 	defer reg.UnregisterClient(clientID)
 
 	if useGitHubCopilotResponsesEndpoint(sdktranslator.FromString("openai"), "gpt-5.4") {
 		t.Fatal("expected dynamic registry definition to take precedence over static fallback")
 	}
+
+	if useGitHubCopilotResponsesEndpoint(sdktranslator.FromString("openai"), "gpt-5.4-mini") {
+		t.Fatal("expected dynamic registry definition to take precedence over static fallback")
+	}
 }
 
 func TestUseGitHubCopilotResponsesEndpoint_DefaultChat(t *testing.T) {
@@ -238,14 +254,14 @@ func TestTranslateGitHubCopilotResponsesNonStreamToClaude_TextMapping(t *testing
 	t.Parallel()
 	resp := []byte(`{"id":"resp_1","model":"gpt-5-codex","output":[{"type":"message","content":[{"type":"output_text","text":"hello"}]}],"usage":{"input_tokens":3,"output_tokens":5}}`)
 	out := translateGitHubCopilotResponsesNonStreamToClaude(resp)
-	if gjson.Get(out, "type").String() != "message" {
-		t.Fatalf("type = %q, want message", gjson.Get(out, "type").String())
+	if gjson.GetBytes(out, "type").String() != "message" {
+		t.Fatalf("type = %q, want message", gjson.GetBytes(out, "type").String())
 	}
-	if gjson.Get(out, "content.0.type").String() != "text" {
-		t.Fatalf("content.0.type = %q, want text", gjson.Get(out, "content.0.type").String())
+	if gjson.GetBytes(out, "content.0.type").String() != "text" {
+		t.Fatalf("content.0.type = %q, want text", gjson.GetBytes(out, "content.0.type").String())
 	}
-	if gjson.Get(out, "content.0.text").String() != "hello" {
-		t.Fatalf("content.0.text = %q, want hello", gjson.Get(out, "content.0.text").String())
+	if gjson.GetBytes(out, "content.0.text").String() != "hello" {
+		t.Fatalf("content.0.text = %q, want hello", gjson.GetBytes(out, "content.0.text").String())
 	}
 }
 
@@ -253,14 +269,14 @@ func TestTranslateGitHubCopilotResponsesNonStreamToClaude_ToolUseMapping(t *test
 	t.Parallel()
 	resp := []byte(`{"id":"resp_2","model":"gpt-5-codex","output":[{"type":"function_call","id":"fc_1","call_id":"call_1","name":"sum","arguments":"{\"a\":1}"}],"usage":{"input_tokens":1,"output_tokens":2}}`)
 	out := translateGitHubCopilotResponsesNonStreamToClaude(resp)
-	if gjson.Get(out, "content.0.type").String() != "tool_use" {
-		t.Fatalf("content.0.type = %q, want tool_use", gjson.Get(out, "content.0.type").String())
+	if gjson.GetBytes(out, "content.0.type").String() != "tool_use" {
+		t.Fatalf("content.0.type = %q, want tool_use", gjson.GetBytes(out, "content.0.type").String())
 	}
-	if gjson.Get(out, "content.0.name").String() != "sum" {
-		t.Fatalf("content.0.name = %q, want sum", gjson.Get(out, "content.0.name").String())
+	if gjson.GetBytes(out, "content.0.name").String() != "sum" {
+		t.Fatalf("content.0.name = %q, want sum", gjson.GetBytes(out, "content.0.name").String())
 	}
-	if gjson.Get(out, "stop_reason").String() != "tool_use" {
-		t.Fatalf("stop_reason = %q, want tool_use", gjson.Get(out, "stop_reason").String())
+	if gjson.GetBytes(out, "stop_reason").String() != "tool_use" {
+		t.Fatalf("stop_reason = %q, want tool_use", gjson.GetBytes(out, "stop_reason").String())
 	}
 }
 
@@ -269,18 +285,24 @@ func TestTranslateGitHubCopilotResponsesStreamToClaude_TextLifecycle(t *testing.
 	var param any
 
 	created := translateGitHubCopilotResponsesStreamToClaude([]byte(`data: {"type":"response.created","response":{"id":"resp_1","model":"gpt-5-codex"}}`), &param)
-	if len(created) == 0 || !strings.Contains(created[0], "message_start") {
+	if len(created) == 0 || !strings.Contains(string(created[0]), "message_start") {
 		t.Fatalf("created events = %#v, want message_start", created)
 	}
 
 	delta := translateGitHubCopilotResponsesStreamToClaude([]byte(`data: {"type":"response.output_text.delta","delta":"he"}`), &param)
-	joinedDelta := strings.Join(delta, "")
+	var joinedDelta string
+	for _, d := range delta {
+		joinedDelta += string(d)
+	}
 	if !strings.Contains(joinedDelta, "content_block_start") || !strings.Contains(joinedDelta, "text_delta") {
 		t.Fatalf("delta events = %#v, want content_block_start + text_delta", delta)
 	}
 
 	completed := translateGitHubCopilotResponsesStreamToClaude([]byte(`data: {"type":"response.completed","response":{"usage":{"input_tokens":7,"output_tokens":9}}}`), &param)
-	joinedCompleted := strings.Join(completed, "")
+	var joinedCompleted string
+	for _, c := range completed {
+		joinedCompleted += string(c)
+	}
 	if !strings.Contains(joinedCompleted, "message_delta") || !strings.Contains(joinedCompleted, "message_stop") {
 		t.Fatalf("completed events = %#v, want message_delta + message_stop", completed)
 	}
@@ -299,15 +321,17 @@ func TestApplyHeaders_XInitiator_UserOnly(t *testing.T) {
 	}
 }
 
-func TestApplyHeaders_XInitiator_UserWhenLastRoleIsUser(t *testing.T) {
+func TestApplyHeaders_XInitiator_AgentWhenLastUserButHistoryHasAssistant(t *testing.T) {
 	t.Parallel()
 	e := &GitHubCopilotExecutor{}
 	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
-	// Last role governs the initiator decision.
-	body := []byte(`{"messages":[{"role":"user","content":"hello"},{"role":"assistant","content":"I will read the file"},{"role":"user","content":"tool result here"}]}`)
+	// When the last role is "user" and the message contains tool_result content,
+	// the request is a continuation (e.g. Claude tool result translated to a
+	// synthetic user message). Should be "agent".
+	body := []byte(`{"messages":[{"role":"user","content":"hello"},{"role":"assistant","content":"I will read the file"},{"role":"user","content":[{"type":"tool_result","tool_use_id":"tu1","content":"file contents..."}]}]}`)
 	e.applyHeaders(req, "token", body)
-	if got := req.Header.Get("X-Initiator"); got != "user" {
-		t.Fatalf("X-Initiator = %q, want user (last role is user)", got)
+	if got := req.Header.Get("X-Initiator"); got != "agent" {
+		t.Fatalf("X-Initiator = %q, want agent (last user contains tool_result)", got)
 	}
 }
 
@@ -315,10 +339,11 @@ func TestApplyHeaders_XInitiator_AgentWithToolRole(t *testing.T) {
 	t.Parallel()
 	e := &GitHubCopilotExecutor{}
 	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
+	// When the last message has role "tool", it's clearly agent-initiated.
 	body := []byte(`{"messages":[{"role":"user","content":"hello"},{"role":"tool","content":"result"}]}`)
 	e.applyHeaders(req, "token", body)
 	if got := req.Header.Get("X-Initiator"); got != "agent" {
-		t.Fatalf("X-Initiator = %q, want agent (tool role exists)", got)
+		t.Fatalf("X-Initiator = %q, want agent (last role is tool)", got)
 	}
 }
 
@@ -333,14 +358,15 @@ func TestApplyHeaders_XInitiator_InputArrayLastAssistantMessage(t *testing.T) {
 	}
 }
 
-func TestApplyHeaders_XInitiator_InputArrayLastUserMessage(t *testing.T) {
+func TestApplyHeaders_XInitiator_InputArrayAgentWhenLastUserButHistoryHasAssistant(t *testing.T) {
 	t.Parallel()
 	e := &GitHubCopilotExecutor{}
 	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
+	// Responses API: last item is user-role but history contains assistant → agent.
 	body := []byte(`{"input":[{"type":"message","role":"assistant","content":[{"type":"output_text","text":"I can help"}]},{"type":"message","role":"user","content":[{"type":"input_text","text":"Do X"}]}]}`)
 	e.applyHeaders(req, "token", body)
-	if got := req.Header.Get("X-Initiator"); got != "user" {
-		t.Fatalf("X-Initiator = %q, want user (last role is user)", got)
+	if got := req.Header.Get("X-Initiator"); got != "agent" {
+		t.Fatalf("X-Initiator = %q, want agent (history has assistant)", got)
 	}
 }
 
@@ -355,6 +381,33 @@ func TestApplyHeaders_XInitiator_InputArrayLastFunctionCallOutput(t *testing.T)
 	}
 }
 
+func TestApplyHeaders_XInitiator_UserInMultiTurnNoTools(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
+	// Genuine multi-turn: user → assistant (plain text) → user follow-up.
+	// No tool messages → should be "user" (not a false-positive).
+	body := []byte(`{"messages":[{"role":"user","content":"hello"},{"role":"assistant","content":"Hi there!"},{"role":"user","content":"what is 2+2?"}]}`)
+	e.applyHeaders(req, "token", body)
+	if got := req.Header.Get("X-Initiator"); got != "user" {
+		t.Fatalf("X-Initiator = %q, want user (genuine multi-turn, no tools)", got)
+	}
+}
+
+func TestApplyHeaders_XInitiator_UserFollowUpAfterToolHistory(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
+	// User follow-up after a completed tool-use conversation.
+	// The last message is a genuine user question — should be "user", not "agent".
+	// This aligns with opencode's behavior: only active tool loops are agent-initiated.
+	body := []byte(`{"messages":[{"role":"user","content":"hello"},{"role":"assistant","content":[{"type":"tool_use","id":"tu1","name":"Read","input":{}}]},{"role":"tool","tool_call_id":"tu1","content":"file data"},{"role":"assistant","content":"I read the file."},{"role":"user","content":"What did we do so far?"}]}`)
+	e.applyHeaders(req, "token", body)
+	if got := req.Header.Get("X-Initiator"); got != "user" {
+		t.Fatalf("X-Initiator = %q, want user (genuine follow-up after tool history)", got)
+	}
+}
+
 // --- Tests for x-github-api-version header (Problem M) ---
 
 func TestApplyHeaders_GitHubAPIVersion(t *testing.T) {
@@ -401,3 +454,364 @@ func TestDetectVisionContent_NoMessages(t *testing.T) {
 		t.Fatal("expected no vision content when messages field is absent")
 	}
 }
+
+// --- Tests for applyGitHubCopilotResponsesDefaults ---
+
+func TestApplyGitHubCopilotResponsesDefaults_SetsAllDefaults(t *testing.T) {
+	t.Parallel()
+	body := []byte(`{"input":"hello","reasoning":{"effort":"medium"}}`)
+	got := applyGitHubCopilotResponsesDefaults(body)
+
+	if gjson.GetBytes(got, "store").Bool() != false {
+		t.Fatalf("store = %v, want false", gjson.GetBytes(got, "store").Raw)
+	}
+	inc := gjson.GetBytes(got, "include")
+	if !inc.IsArray() || inc.Array()[0].String() != "reasoning.encrypted_content" {
+		t.Fatalf("include = %s, want [\"reasoning.encrypted_content\"]", inc.Raw)
+	}
+	if gjson.GetBytes(got, "reasoning.summary").String() != "auto" {
+		t.Fatalf("reasoning.summary = %q, want auto", gjson.GetBytes(got, "reasoning.summary").String())
+	}
+}
+
+func TestApplyGitHubCopilotResponsesDefaults_DoesNotOverrideExisting(t *testing.T) {
+	t.Parallel()
+	body := []byte(`{"input":"hello","store":true,"include":["other"],"reasoning":{"effort":"high","summary":"concise"}}`)
+	got := applyGitHubCopilotResponsesDefaults(body)
+
+	if gjson.GetBytes(got, "store").Bool() != true {
+		t.Fatalf("store should not be overridden, got %s", gjson.GetBytes(got, "store").Raw)
+	}
+	if gjson.GetBytes(got, "include").Array()[0].String() != "other" {
+		t.Fatalf("include should not be overridden, got %s", gjson.GetBytes(got, "include").Raw)
+	}
+	if gjson.GetBytes(got, "reasoning.summary").String() != "concise" {
+		t.Fatalf("reasoning.summary should not be overridden, got %q", gjson.GetBytes(got, "reasoning.summary").String())
+	}
+}
+
+func TestApplyGitHubCopilotResponsesDefaults_NoReasoningEffort(t *testing.T) {
+	t.Parallel()
+	body := []byte(`{"input":"hello"}`)
+	got := applyGitHubCopilotResponsesDefaults(body)
+
+	if gjson.GetBytes(got, "store").Bool() != false {
+		t.Fatalf("store = %v, want false", gjson.GetBytes(got, "store").Raw)
+	}
+	// reasoning.summary should NOT be set when reasoning.effort is absent
+	if gjson.GetBytes(got, "reasoning.summary").Exists() {
+		t.Fatalf("reasoning.summary should not be set when reasoning.effort is absent, got %q", gjson.GetBytes(got, "reasoning.summary").String())
+	}
+}
+
+// --- Tests for normalizeGitHubCopilotReasoningField ---
+
+func TestNormalizeReasoningField_NonStreaming(t *testing.T) {
+	t.Parallel()
+	data := []byte(`{"choices":[{"message":{"content":"hello","reasoning_text":"I think..."}}]}`)
+	got := normalizeGitHubCopilotReasoningField(data)
+	rc := gjson.GetBytes(got, "choices.0.message.reasoning_content").String()
+	if rc != "I think..." {
+		t.Fatalf("reasoning_content = %q, want %q", rc, "I think...")
+	}
+}
+
+func TestNormalizeReasoningField_Streaming(t *testing.T) {
+	t.Parallel()
+	data := []byte(`{"choices":[{"delta":{"reasoning_text":"thinking delta"}}]}`)
+	got := normalizeGitHubCopilotReasoningField(data)
+	rc := gjson.GetBytes(got, "choices.0.delta.reasoning_content").String()
+	if rc != "thinking delta" {
+		t.Fatalf("reasoning_content = %q, want %q", rc, "thinking delta")
+	}
+}
+
+func TestNormalizeReasoningField_PreservesExistingReasoningContent(t *testing.T) {
+	t.Parallel()
+	data := []byte(`{"choices":[{"message":{"reasoning_text":"old","reasoning_content":"existing"}}]}`)
+	got := normalizeGitHubCopilotReasoningField(data)
+	rc := gjson.GetBytes(got, "choices.0.message.reasoning_content").String()
+	if rc != "existing" {
+		t.Fatalf("reasoning_content = %q, want %q (should not overwrite)", rc, "existing")
+	}
+}
+
+func TestNormalizeReasoningField_MultiChoice(t *testing.T) {
+	t.Parallel()
+	data := []byte(`{"choices":[{"message":{"reasoning_text":"thought-0"}},{"message":{"reasoning_text":"thought-1"}}]}`)
+	got := normalizeGitHubCopilotReasoningField(data)
+	rc0 := gjson.GetBytes(got, "choices.0.message.reasoning_content").String()
+	rc1 := gjson.GetBytes(got, "choices.1.message.reasoning_content").String()
+	if rc0 != "thought-0" {
+		t.Fatalf("choices[0].reasoning_content = %q, want %q", rc0, "thought-0")
+	}
+	if rc1 != "thought-1" {
+		t.Fatalf("choices[1].reasoning_content = %q, want %q", rc1, "thought-1")
+	}
+}
+
+func TestNormalizeReasoningField_NoChoices(t *testing.T) {
+	t.Parallel()
+	data := []byte(`{"id":"chatcmpl-123"}`)
+	got := normalizeGitHubCopilotReasoningField(data)
+	if string(got) != string(data) {
+		t.Fatalf("expected no change, got %s", string(got))
+	}
+}
+
+func TestApplyHeaders_OpenAIIntentValue(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	req, _ := http.NewRequest(http.MethodPost, "https://example.com", nil)
+	e.applyHeaders(req, "token", nil)
+	if got := req.Header.Get("Openai-Intent"); got != "conversation-edits" {
+		t.Fatalf("Openai-Intent = %q, want conversation-edits", got)
+	}
+}
+
+// --- Tests for CountTokens (local tiktoken estimation) ---
+
+func TestCountTokens_ReturnsPositiveCount(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	body := []byte(`{"model":"gpt-4o","messages":[{"role":"user","content":"Hello, world!"}]}`)
+	resp, err := e.CountTokens(context.Background(), nil, cliproxyexecutor.Request{
+		Model:   "gpt-4o",
+		Payload: body,
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai"),
+	})
+	if err != nil {
+		t.Fatalf("CountTokens() error: %v", err)
+	}
+	if len(resp.Payload) == 0 {
+		t.Fatal("CountTokens() returned empty payload")
+	}
+	// The response should contain a positive token count.
+	tokens := gjson.GetBytes(resp.Payload, "usage.prompt_tokens").Int()
+	if tokens <= 0 {
+		t.Fatalf("expected positive token count, got %d", tokens)
+	}
+}
+
+func TestCountTokens_ClaudeSourceFormatTranslates(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	body := []byte(`{"model":"claude-sonnet-4","messages":[{"role":"user","content":"Tell me a joke"}],"max_tokens":1024}`)
+	resp, err := e.CountTokens(context.Background(), nil, cliproxyexecutor.Request{
+		Model:   "claude-sonnet-4",
+		Payload: body,
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("claude"),
+	})
+	if err != nil {
+		t.Fatalf("CountTokens() error: %v", err)
+	}
+	// Claude source format → should get input_tokens in response
+	inputTokens := gjson.GetBytes(resp.Payload, "input_tokens").Int()
+	if inputTokens <= 0 {
+		// Fallback: check usage.prompt_tokens (depends on translator registration)
+		promptTokens := gjson.GetBytes(resp.Payload, "usage.prompt_tokens").Int()
+		if promptTokens <= 0 {
+			t.Fatalf("expected positive token count, got payload: %s", resp.Payload)
+		}
+	}
+}
+
+func TestCountTokens_EmptyPayload(t *testing.T) {
+	t.Parallel()
+	e := &GitHubCopilotExecutor{}
+	resp, err := e.CountTokens(context.Background(), nil, cliproxyexecutor.Request{
+		Model:   "gpt-4o",
+		Payload: []byte(`{"model":"gpt-4o","messages":[]}`),
+	}, cliproxyexecutor.Options{
+		SourceFormat: sdktranslator.FromString("openai"),
+	})
+	if err != nil {
+		t.Fatalf("CountTokens() error: %v", err)
+	}
+	tokens := gjson.GetBytes(resp.Payload, "usage.prompt_tokens").Int()
+	// Empty messages should return 0 tokens.
+	if tokens != 0 {
+		t.Fatalf("expected 0 tokens for empty messages, got %d", tokens)
+	}
+}
+
+func TestStripUnsupportedBetas_RemovesContext1M(t *testing.T) {
+	t.Parallel()
+
+	body := []byte(`{"model":"claude-opus-4.6","betas":["interleaved-thinking-2025-05-14","context-1m-2025-08-07","claude-code-20250219"],"messages":[]}`)
+	result := stripUnsupportedBetas(body)
+
+	betas := gjson.GetBytes(result, "betas")
+	if !betas.Exists() {
+		t.Fatal("betas field should still exist after stripping")
+	}
+	for _, item := range betas.Array() {
+		if item.String() == "context-1m-2025-08-07" {
+			t.Fatal("context-1m-2025-08-07 should have been stripped")
+		}
+	}
+	// Other betas should be preserved
+	found := false
+	for _, item := range betas.Array() {
+		if item.String() == "interleaved-thinking-2025-05-14" {
+			found = true
+		}
+	}
+	if !found {
+		t.Fatal("other betas should be preserved")
+	}
+}
+
+func TestStripUnsupportedBetas_NoBetasField(t *testing.T) {
+	t.Parallel()
+
+	body := []byte(`{"model":"gpt-4o","messages":[]}`)
+	result := stripUnsupportedBetas(body)
+
+	// Should be unchanged
+	if string(result) != string(body) {
+		t.Fatalf("body should be unchanged when no betas field exists, got %s", string(result))
+	}
+}
+
+func TestStripUnsupportedBetas_MetadataBetas(t *testing.T) {
+	t.Parallel()
+
+	body := []byte(`{"model":"claude-opus-4.6","metadata":{"betas":["context-1m-2025-08-07","other-beta"]},"messages":[]}`)
+	result := stripUnsupportedBetas(body)
+
+	betas := gjson.GetBytes(result, "metadata.betas")
+	if !betas.Exists() {
+		t.Fatal("metadata.betas field should still exist after stripping")
+	}
+	for _, item := range betas.Array() {
+		if item.String() == "context-1m-2025-08-07" {
+			t.Fatal("context-1m-2025-08-07 should have been stripped from metadata.betas")
+		}
+	}
+	if betas.Array()[0].String() != "other-beta" {
+		t.Fatal("other betas in metadata.betas should be preserved")
+	}
+}
+
+func TestStripUnsupportedBetas_AllBetasStripped(t *testing.T) {
+	t.Parallel()
+
+	body := []byte(`{"model":"claude-opus-4.6","betas":["context-1m-2025-08-07"],"messages":[]}`)
+	result := stripUnsupportedBetas(body)
+
+	betas := gjson.GetBytes(result, "betas")
+	if betas.Exists() {
+		t.Fatal("betas field should be deleted when all betas are stripped")
+	}
+}
+
+func TestCopilotModelEntry_Limits(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name         string
+		capabilities map[string]any
+		wantNil      bool
+		wantPrompt   int
+		wantOutput   int
+		wantContext  int
+	}{
+		{
+			name:         "nil capabilities",
+			capabilities: nil,
+			wantNil:      true,
+		},
+		{
+			name:         "no limits key",
+			capabilities: map[string]any{"family": "claude-opus-4.6"},
+			wantNil:      true,
+		},
+		{
+			name:         "limits is not a map",
+			capabilities: map[string]any{"limits": "invalid"},
+			wantNil:      true,
+		},
+		{
+			name: "all zero values",
+			capabilities: map[string]any{
+				"limits": map[string]any{
+					"max_context_window_tokens": float64(0),
+					"max_prompt_tokens":         float64(0),
+					"max_output_tokens":         float64(0),
+				},
+			},
+			wantNil: true,
+		},
+		{
+			name: "individual account limits (128K prompt)",
+			capabilities: map[string]any{
+				"limits": map[string]any{
+					"max_context_window_tokens": float64(144000),
+					"max_prompt_tokens":         float64(128000),
+					"max_output_tokens":         float64(64000),
+				},
+			},
+			wantNil:     false,
+			wantPrompt:  128000,
+			wantOutput:  64000,
+			wantContext: 144000,
+		},
+		{
+			name: "business account limits (168K prompt)",
+			capabilities: map[string]any{
+				"limits": map[string]any{
+					"max_context_window_tokens": float64(200000),
+					"max_prompt_tokens":         float64(168000),
+					"max_output_tokens":         float64(32000),
+				},
+			},
+			wantNil:     false,
+			wantPrompt:  168000,
+			wantOutput:  32000,
+			wantContext: 200000,
+		},
+		{
+			name: "partial limits (only prompt)",
+			capabilities: map[string]any{
+				"limits": map[string]any{
+					"max_prompt_tokens": float64(128000),
+				},
+			},
+			wantNil:    false,
+			wantPrompt: 128000,
+			wantOutput: 0,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+			entry := copilotauth.CopilotModelEntry{
+				ID:           "claude-opus-4.6",
+				Capabilities: tt.capabilities,
+			}
+			limits := entry.Limits()
+			if tt.wantNil {
+				if limits != nil {
+					t.Fatalf("expected nil limits, got %+v", limits)
+				}
+				return
+			}
+			if limits == nil {
+				t.Fatal("expected non-nil limits, got nil")
+			}
+			if limits.MaxPromptTokens != tt.wantPrompt {
+				t.Errorf("MaxPromptTokens = %d, want %d", limits.MaxPromptTokens, tt.wantPrompt)
+			}
+			if limits.MaxOutputTokens != tt.wantOutput {
+				t.Errorf("MaxOutputTokens = %d, want %d", limits.MaxOutputTokens, tt.wantOutput)
+			}
+			if tt.wantContext > 0 && limits.MaxContextWindowTokens != tt.wantContext {
+				t.Errorf("MaxContextWindowTokens = %d, want %d", limits.MaxContextWindowTokens, tt.wantContext)
+			}
+		})
+	}
+}

internal/runtime/executor/helps/cache_helpers.go

@@ -1,11 +1,11 @@
-package executor
+package helps
 
 import (
 	"sync"
 	"time"
 )
 
-type codexCache struct {
+type CodexCache struct {
 	ID     string
 	Expire time.Time
 }
@@ -13,7 +13,7 @@ type codexCache struct {
 // codexCacheMap stores prompt cache IDs keyed by model+user_id.
 // Protected by codexCacheMu. Entries expire after 1 hour.
 var (
-	codexCacheMap = make(map[string]codexCache)
+	codexCacheMap = make(map[string]CodexCache)
 	codexCacheMu  sync.RWMutex
 )
 
@@ -50,20 +50,20 @@ func purgeExpiredCodexCache() {
 	}
 }
 
-// getCodexCache retrieves a cached entry, returning ok=false if not found or expired.
-func getCodexCache(key string) (codexCache, bool) {
+// GetCodexCache retrieves a cached entry, returning ok=false if not found or expired.
+func GetCodexCache(key string) (CodexCache, bool) {
 	codexCacheCleanupOnce.Do(startCodexCacheCleanup)
 	codexCacheMu.RLock()
 	cache, ok := codexCacheMap[key]
 	codexCacheMu.RUnlock()
 	if !ok || cache.Expire.Before(time.Now()) {
-		return codexCache{}, false
+		return CodexCache{}, false
 	}
 	return cache, true
 }
 
-// setCodexCache stores a cache entry.
-func setCodexCache(key string, cache codexCache) {
+// SetCodexCache stores a cache entry.
+func SetCodexCache(key string, cache CodexCache) {
 	codexCacheCleanupOnce.Do(startCodexCacheCleanup)
 	codexCacheMu.Lock()
 	codexCacheMap[key] = cache

internal/runtime/executor/helps/claude_device_profile.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"crypto/sha256"
@@ -32,7 +32,7 @@ var (
 	claudeDeviceProfileCacheMu          sync.RWMutex
 	claudeDeviceProfileCacheCleanupOnce sync.Once
 
-	claudeDeviceProfileBeforeCandidateStore func(claudeDeviceProfile)
+	ClaudeDeviceProfileBeforeCandidateStore func(ClaudeDeviceProfile)
 )
 
 type claudeCLIVersion struct {
@@ -63,29 +63,43 @@ func (v claudeCLIVersion) Compare(other claudeCLIVersion) int {
 	}
 }
 
-type claudeDeviceProfile struct {
+type ClaudeDeviceProfile struct {
 	UserAgent      string
 	PackageVersion string
 	RuntimeVersion string
 	OS             string
 	Arch           string
-	Version        claudeCLIVersion
-	HasVersion     bool
+	version        claudeCLIVersion
+	hasVersion     bool
 }
 
 type claudeDeviceProfileCacheEntry struct {
-	profile claudeDeviceProfile
+	profile ClaudeDeviceProfile
 	expire  time.Time
 }
 
-func claudeDeviceProfileStabilizationEnabled(cfg *config.Config) bool {
+func ClaudeDeviceProfileStabilizationEnabled(cfg *config.Config) bool {
 	if cfg == nil || cfg.ClaudeHeaderDefaults.StabilizeDeviceProfile == nil {
 		return false
 	}
 	return *cfg.ClaudeHeaderDefaults.StabilizeDeviceProfile
 }
 
-func defaultClaudeDeviceProfile(cfg *config.Config) claudeDeviceProfile {
+func ResetClaudeDeviceProfileCache() {
+	claudeDeviceProfileCacheMu.Lock()
+	claudeDeviceProfileCache = make(map[string]claudeDeviceProfileCacheEntry)
+	claudeDeviceProfileCacheMu.Unlock()
+}
+
+func MapStainlessOS() string {
+	return mapStainlessOS()
+}
+
+func MapStainlessArch() string {
+	return mapStainlessArch()
+}
+
+func defaultClaudeDeviceProfile(cfg *config.Config) ClaudeDeviceProfile {
 	hdrDefault := func(cfgVal, fallback string) string {
 		if strings.TrimSpace(cfgVal) != "" {
 			return strings.TrimSpace(cfgVal)
@@ -98,7 +112,7 @@ func defaultClaudeDeviceProfile(cfg *config.Config) claudeDeviceProfile {
 		hd = cfg.ClaudeHeaderDefaults
 	}
 
-	profile := claudeDeviceProfile{
+	profile := ClaudeDeviceProfile{
 		UserAgent:      hdrDefault(hd.UserAgent, defaultClaudeFingerprintUserAgent),
 		PackageVersion: hdrDefault(hd.PackageVersion, defaultClaudeFingerprintPackageVersion),
 		RuntimeVersion: hdrDefault(hd.RuntimeVersion, defaultClaudeFingerprintRuntimeVersion),
@@ -106,8 +120,8 @@ func defaultClaudeDeviceProfile(cfg *config.Config) claudeDeviceProfile {
 		Arch:           hdrDefault(hd.Arch, defaultClaudeFingerprintArch),
 	}
 	if version, ok := parseClaudeCLIVersion(profile.UserAgent); ok {
-		profile.Version = version
-		profile.HasVersion = true
+		profile.version = version
+		profile.hasVersion = true
 	}
 	return profile
 }
@@ -162,17 +176,17 @@ func parseClaudeCLIVersion(userAgent string) (claudeCLIVersion, bool) {
 	return claudeCLIVersion{major: major, minor: minor, patch: patch}, true
 }
 
-func shouldUpgradeClaudeDeviceProfile(candidate, current claudeDeviceProfile) bool {
-	if candidate.UserAgent == "" || !candidate.HasVersion {
+func shouldUpgradeClaudeDeviceProfile(candidate, current ClaudeDeviceProfile) bool {
+	if candidate.UserAgent == "" || !candidate.hasVersion {
 		return false
 	}
-	if current.UserAgent == "" || !current.HasVersion {
+	if current.UserAgent == "" || !current.hasVersion {
 		return true
 	}
-	return candidate.Version.Compare(current.Version) > 0
+	return candidate.version.Compare(current.version) > 0
 }
 
-func pinClaudeDeviceProfilePlatform(profile, baseline claudeDeviceProfile) claudeDeviceProfile {
+func pinClaudeDeviceProfilePlatform(profile, baseline ClaudeDeviceProfile) ClaudeDeviceProfile {
 	profile.OS = baseline.OS
 	profile.Arch = baseline.Arch
 	return profile
@@ -180,38 +194,38 @@ func pinClaudeDeviceProfilePlatform(profile, baseline claudeDeviceProfile) claud
 
 // normalizeClaudeDeviceProfile keeps stabilized profiles pinned to the current
 // baseline platform and enforces the baseline software fingerprint as a floor.
-func normalizeClaudeDeviceProfile(profile, baseline claudeDeviceProfile) claudeDeviceProfile {
+func normalizeClaudeDeviceProfile(profile, baseline ClaudeDeviceProfile) ClaudeDeviceProfile {
 	profile = pinClaudeDeviceProfilePlatform(profile, baseline)
-	if profile.UserAgent == "" || !profile.HasVersion || shouldUpgradeClaudeDeviceProfile(baseline, profile) {
+	if profile.UserAgent == "" || !profile.hasVersion || shouldUpgradeClaudeDeviceProfile(baseline, profile) {
 		profile.UserAgent = baseline.UserAgent
 		profile.PackageVersion = baseline.PackageVersion
 		profile.RuntimeVersion = baseline.RuntimeVersion
-		profile.Version = baseline.Version
-		profile.HasVersion = baseline.HasVersion
+		profile.version = baseline.version
+		profile.hasVersion = baseline.hasVersion
 	}
 	return profile
 }
 
-func extractClaudeDeviceProfile(headers http.Header, cfg *config.Config) (claudeDeviceProfile, bool) {
+func extractClaudeDeviceProfile(headers http.Header, cfg *config.Config) (ClaudeDeviceProfile, bool) {
 	if headers == nil {
-		return claudeDeviceProfile{}, false
+		return ClaudeDeviceProfile{}, false
 	}
 
 	userAgent := strings.TrimSpace(headers.Get("User-Agent"))
 	version, ok := parseClaudeCLIVersion(userAgent)
 	if !ok {
-		return claudeDeviceProfile{}, false
+		return ClaudeDeviceProfile{}, false
 	}
 
 	baseline := defaultClaudeDeviceProfile(cfg)
-	profile := claudeDeviceProfile{
+	profile := ClaudeDeviceProfile{
 		UserAgent:      userAgent,
 		PackageVersion: firstNonEmptyHeader(headers, "X-Stainless-Package-Version", baseline.PackageVersion),
 		RuntimeVersion: firstNonEmptyHeader(headers, "X-Stainless-Runtime-Version", baseline.RuntimeVersion),
 		OS:             firstNonEmptyHeader(headers, "X-Stainless-Os", baseline.OS),
 		Arch:           firstNonEmptyHeader(headers, "X-Stainless-Arch", baseline.Arch),
-		Version:        version,
-		HasVersion:     true,
+		version:        version,
+		hasVersion:     true,
 	}
 	return profile, true
 }
@@ -263,7 +277,7 @@ func purgeExpiredClaudeDeviceProfiles() {
 	claudeDeviceProfileCacheMu.Unlock()
 }
 
-func resolveClaudeDeviceProfile(auth *cliproxyauth.Auth, apiKey string, headers http.Header, cfg *config.Config) claudeDeviceProfile {
+func ResolveClaudeDeviceProfile(auth *cliproxyauth.Auth, apiKey string, headers http.Header, cfg *config.Config) ClaudeDeviceProfile {
 	claudeDeviceProfileCacheCleanupOnce.Do(startClaudeDeviceProfileCacheCleanup)
 
 	cacheKey := claudeDeviceProfileCacheKey(auth, apiKey)
@@ -283,8 +297,8 @@ func resolveClaudeDeviceProfile(auth *cliproxyauth.Auth, apiKey string, headers
 	claudeDeviceProfileCacheMu.RUnlock()
 
 	if hasCandidate {
-		if claudeDeviceProfileBeforeCandidateStore != nil {
-			claudeDeviceProfileBeforeCandidateStore(candidate)
+		if ClaudeDeviceProfileBeforeCandidateStore != nil {
+			ClaudeDeviceProfileBeforeCandidateStore(candidate)
 		}
 
 		claudeDeviceProfileCacheMu.Lock()
@@ -324,7 +338,7 @@ func resolveClaudeDeviceProfile(auth *cliproxyauth.Auth, apiKey string, headers
 	return baseline
 }
 
-func applyClaudeDeviceProfileHeaders(r *http.Request, profile claudeDeviceProfile) {
+func ApplyClaudeDeviceProfileHeaders(r *http.Request, profile ClaudeDeviceProfile) {
 	if r == nil {
 		return
 	}
@@ -344,7 +358,17 @@ func applyClaudeDeviceProfileHeaders(r *http.Request, profile claudeDeviceProfil
 	r.Header.Set("X-Stainless-Arch", profile.Arch)
 }
 
-func applyClaudeLegacyDeviceHeaders(r *http.Request, ginHeaders http.Header, cfg *config.Config) {
+// DefaultClaudeVersion returns the version string (e.g. "2.1.63") from the
+// current baseline device profile. It extracts the version from the User-Agent.
+func DefaultClaudeVersion(cfg *config.Config) string {
+	profile := defaultClaudeDeviceProfile(cfg)
+	if version, ok := parseClaudeCLIVersion(profile.UserAgent); ok {
+		return strconv.Itoa(version.major) + "." + strconv.Itoa(version.minor) + "." + strconv.Itoa(version.patch)
+	}
+	return "2.1.63"
+}
+
+func ApplyClaudeLegacyDeviceHeaders(r *http.Request, ginHeaders http.Header, cfg *config.Config) {
 	if r == nil {
 		return
 	}

internal/runtime/executor/helps/cloak_obfuscate.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"regexp"
@@ -18,9 +18,9 @@ type SensitiveWordMatcher struct {
 	regex *regexp.Regexp
 }
 
-// buildSensitiveWordMatcher compiles a regex from the word list.
+// BuildSensitiveWordMatcher compiles a regex from the word list.
 // Words are sorted by length (longest first) for proper matching.
-func buildSensitiveWordMatcher(words []string) *SensitiveWordMatcher {
+func BuildSensitiveWordMatcher(words []string) *SensitiveWordMatcher {
 	if len(words) == 0 {
 		return nil
 	}
@@ -81,9 +81,9 @@ func (m *SensitiveWordMatcher) obfuscateText(text string) string {
 	return m.regex.ReplaceAllStringFunc(text, obfuscateWord)
 }
 
-// obfuscateSensitiveWords processes the payload and obfuscates sensitive words
+// ObfuscateSensitiveWords processes the payload and obfuscates sensitive words
 // in system blocks and message content.
-func obfuscateSensitiveWords(payload []byte, matcher *SensitiveWordMatcher) []byte {
+func ObfuscateSensitiveWords(payload []byte, matcher *SensitiveWordMatcher) []byte {
 	if matcher == nil || matcher.regex == nil {
 		return payload
 	}

internal/runtime/executor/helps/cloak_utils.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"crypto/rand"
@@ -28,9 +28,17 @@ func isValidUserID(userID string) bool {
 	return userIDPattern.MatchString(userID)
 }
 
-// shouldCloak determines if request should be cloaked based on config and client User-Agent.
+func GenerateFakeUserID() string {
+	return generateFakeUserID()
+}
+
+func IsValidUserID(userID string) bool {
+	return isValidUserID(userID)
+}
+
+// ShouldCloak determines if request should be cloaked based on config and client User-Agent.
 // Returns true if cloaking should be applied.
-func shouldCloak(cloakMode string, userAgent string) bool {
+func ShouldCloak(cloakMode string, userAgent string) bool {
 	switch strings.ToLower(cloakMode) {
 	case "always":
 		return true

internal/runtime/executor/helps/logging_helpers.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"bytes"
@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"html"
 	"net/http"
+	"net/url"
 	"sort"
 	"strings"
 	"time"
@@ -19,13 +20,14 @@ import (
 )
 
 const (
-	apiAttemptsKey = "API_UPSTREAM_ATTEMPTS"
-	apiRequestKey  = "API_REQUEST"
-	apiResponseKey = "API_RESPONSE"
+	apiAttemptsKey          = "API_UPSTREAM_ATTEMPTS"
+	apiRequestKey           = "API_REQUEST"
+	apiResponseKey          = "API_RESPONSE"
+	apiWebsocketTimelineKey = "API_WEBSOCKET_TIMELINE"
 )
 
-// upstreamRequestLog captures the outbound upstream request details for logging.
-type upstreamRequestLog struct {
+// UpstreamRequestLog captures the outbound upstream request details for logging.
+type UpstreamRequestLog struct {
 	URL       string
 	Method    string
 	Headers   http.Header
@@ -46,11 +48,12 @@ type upstreamAttempt struct {
 	headersWritten       bool
 	bodyStarted          bool
 	bodyHasContent       bool
+	prevWasSSEEvent      bool
 	errorWritten         bool
 }
 
-// recordAPIRequest stores the upstream request metadata in Gin context for request logging.
-func recordAPIRequest(ctx context.Context, cfg *config.Config, info upstreamRequestLog) {
+// RecordAPIRequest stores the upstream request metadata in Gin context for request logging.
+func RecordAPIRequest(ctx context.Context, cfg *config.Config, info UpstreamRequestLog) {
 	if cfg == nil || !cfg.RequestLog {
 		return
 	}
@@ -96,8 +99,8 @@ func recordAPIRequest(ctx context.Context, cfg *config.Config, info upstreamRequ
 	updateAggregatedRequest(ginCtx, attempts)
 }
 
-// recordAPIResponseMetadata captures upstream response status/header information for the latest attempt.
-func recordAPIResponseMetadata(ctx context.Context, cfg *config.Config, status int, headers http.Header) {
+// RecordAPIResponseMetadata captures upstream response status/header information for the latest attempt.
+func RecordAPIResponseMetadata(ctx context.Context, cfg *config.Config, status int, headers http.Header) {
 	if cfg == nil || !cfg.RequestLog {
 		return
 	}
@@ -122,8 +125,8 @@ func recordAPIResponseMetadata(ctx context.Context, cfg *config.Config, status i
 	updateAggregatedResponse(ginCtx, attempts)
 }
 
-// recordAPIResponseError adds an error entry for the latest attempt when no HTTP response is available.
-func recordAPIResponseError(ctx context.Context, cfg *config.Config, err error) {
+// RecordAPIResponseError adds an error entry for the latest attempt when no HTTP response is available.
+func RecordAPIResponseError(ctx context.Context, cfg *config.Config, err error) {
 	if cfg == nil || !cfg.RequestLog || err == nil {
 		return
 	}
@@ -147,8 +150,8 @@ func recordAPIResponseError(ctx context.Context, cfg *config.Config, err error)
 	updateAggregatedResponse(ginCtx, attempts)
 }
 
-// appendAPIResponseChunk appends an upstream response chunk to Gin context for request logging.
-func appendAPIResponseChunk(ctx context.Context, cfg *config.Config, chunk []byte) {
+// AppendAPIResponseChunk appends an upstream response chunk to Gin context for request logging.
+func AppendAPIResponseChunk(ctx context.Context, cfg *config.Config, chunk []byte) {
 	if cfg == nil || !cfg.RequestLog {
 		return
 	}
@@ -173,15 +176,157 @@ func appendAPIResponseChunk(ctx context.Context, cfg *config.Config, chunk []byt
 		attempt.response.WriteString("Body:\n")
 		attempt.bodyStarted = true
 	}
+	currentChunkIsSSEEvent := bytes.HasPrefix(data, []byte("event:"))
+	currentChunkIsSSEData := bytes.HasPrefix(data, []byte("data:"))
 	if attempt.bodyHasContent {
-		attempt.response.WriteString("\n\n")
+		separator := "\n\n"
+		if attempt.prevWasSSEEvent && currentChunkIsSSEData {
+			separator = "\n"
+		}
+		attempt.response.WriteString(separator)
 	}
 	attempt.response.WriteString(string(data))
 	attempt.bodyHasContent = true
+	attempt.prevWasSSEEvent = currentChunkIsSSEEvent
 
 	updateAggregatedResponse(ginCtx, attempts)
 }
 
+// RecordAPIWebsocketRequest stores an upstream websocket request event in Gin context.
+func RecordAPIWebsocketRequest(ctx context.Context, cfg *config.Config, info UpstreamRequestLog) {
+	if cfg == nil || !cfg.RequestLog {
+		return
+	}
+	ginCtx := ginContextFrom(ctx)
+	if ginCtx == nil {
+		return
+	}
+
+	builder := &strings.Builder{}
+	builder.WriteString(fmt.Sprintf("Timestamp: %s\n", time.Now().Format(time.RFC3339Nano)))
+	builder.WriteString("Event: api.websocket.request\n")
+	if info.URL != "" {
+		builder.WriteString(fmt.Sprintf("Upstream URL: %s\n", info.URL))
+	}
+	if auth := formatAuthInfo(info); auth != "" {
+		builder.WriteString(fmt.Sprintf("Auth: %s\n", auth))
+	}
+	builder.WriteString("Headers:\n")
+	writeHeaders(builder, info.Headers)
+	builder.WriteString("\nBody:\n")
+	if len(info.Body) > 0 {
+		builder.Write(info.Body)
+	} else {
+		builder.WriteString("<empty>")
+	}
+	builder.WriteString("\n")
+
+	appendAPIWebsocketTimeline(ginCtx, []byte(builder.String()))
+}
+
+// RecordAPIWebsocketHandshake stores the upstream websocket handshake response metadata.
+func RecordAPIWebsocketHandshake(ctx context.Context, cfg *config.Config, status int, headers http.Header) {
+	if cfg == nil || !cfg.RequestLog {
+		return
+	}
+	ginCtx := ginContextFrom(ctx)
+	if ginCtx == nil {
+		return
+	}
+
+	builder := &strings.Builder{}
+	builder.WriteString(fmt.Sprintf("Timestamp: %s\n", time.Now().Format(time.RFC3339Nano)))
+	builder.WriteString("Event: api.websocket.handshake\n")
+	if status > 0 {
+		builder.WriteString(fmt.Sprintf("Status: %d\n", status))
+	}
+	builder.WriteString("Headers:\n")
+	writeHeaders(builder, headers)
+	builder.WriteString("\n")
+
+	appendAPIWebsocketTimeline(ginCtx, []byte(builder.String()))
+}
+
+// RecordAPIWebsocketUpgradeRejection stores a rejected websocket upgrade as an HTTP attempt.
+func RecordAPIWebsocketUpgradeRejection(ctx context.Context, cfg *config.Config, info UpstreamRequestLog, status int, headers http.Header, body []byte) {
+	if cfg == nil || !cfg.RequestLog {
+		return
+	}
+	ginCtx := ginContextFrom(ctx)
+	if ginCtx == nil {
+		return
+	}
+
+	RecordAPIRequest(ctx, cfg, info)
+	RecordAPIResponseMetadata(ctx, cfg, status, headers)
+	AppendAPIResponseChunk(ctx, cfg, body)
+}
+
+// WebsocketUpgradeRequestURL converts a websocket URL back to its HTTP handshake URL for logging.
+func WebsocketUpgradeRequestURL(rawURL string) string {
+	trimmedURL := strings.TrimSpace(rawURL)
+	if trimmedURL == "" {
+		return ""
+	}
+	parsed, err := url.Parse(trimmedURL)
+	if err != nil {
+		return trimmedURL
+	}
+	switch strings.ToLower(parsed.Scheme) {
+	case "ws":
+		parsed.Scheme = "http"
+	case "wss":
+		parsed.Scheme = "https"
+	}
+	return parsed.String()
+}
+
+// AppendAPIWebsocketResponse stores an upstream websocket response frame in Gin context.
+func AppendAPIWebsocketResponse(ctx context.Context, cfg *config.Config, payload []byte) {
+	if cfg == nil || !cfg.RequestLog {
+		return
+	}
+	data := bytes.TrimSpace(payload)
+	if len(data) == 0 {
+		return
+	}
+	ginCtx := ginContextFrom(ctx)
+	if ginCtx == nil {
+		return
+	}
+	markAPIResponseTimestamp(ginCtx)
+
+	builder := &strings.Builder{}
+	builder.WriteString(fmt.Sprintf("Timestamp: %s\n", time.Now().Format(time.RFC3339Nano)))
+	builder.WriteString("Event: api.websocket.response\n")
+	builder.Write(data)
+	builder.WriteString("\n")
+
+	appendAPIWebsocketTimeline(ginCtx, []byte(builder.String()))
+}
+
+// RecordAPIWebsocketError stores an upstream websocket error event in Gin context.
+func RecordAPIWebsocketError(ctx context.Context, cfg *config.Config, stage string, err error) {
+	if cfg == nil || !cfg.RequestLog || err == nil {
+		return
+	}
+	ginCtx := ginContextFrom(ctx)
+	if ginCtx == nil {
+		return
+	}
+	markAPIResponseTimestamp(ginCtx)
+
+	builder := &strings.Builder{}
+	builder.WriteString(fmt.Sprintf("Timestamp: %s\n", time.Now().Format(time.RFC3339Nano)))
+	builder.WriteString("Event: api.websocket.error\n")
+	if trimmed := strings.TrimSpace(stage); trimmed != "" {
+		builder.WriteString(fmt.Sprintf("Stage: %s\n", trimmed))
+	}
+	builder.WriteString(fmt.Sprintf("Error: %s\n", err.Error()))
+
+	appendAPIWebsocketTimeline(ginCtx, []byte(builder.String()))
+}
+
 func ginContextFrom(ctx context.Context) *gin.Context {
 	ginCtx, _ := ctx.Value("gin").(*gin.Context)
 	return ginCtx
@@ -259,6 +404,40 @@ func updateAggregatedResponse(ginCtx *gin.Context, attempts []*upstreamAttempt)
 	ginCtx.Set(apiResponseKey, []byte(builder.String()))
 }
 
+func appendAPIWebsocketTimeline(ginCtx *gin.Context, chunk []byte) {
+	if ginCtx == nil {
+		return
+	}
+	data := bytes.TrimSpace(chunk)
+	if len(data) == 0 {
+		return
+	}
+	if existing, exists := ginCtx.Get(apiWebsocketTimelineKey); exists {
+		if existingBytes, ok := existing.([]byte); ok && len(existingBytes) > 0 {
+			combined := make([]byte, 0, len(existingBytes)+len(data)+2)
+			combined = append(combined, existingBytes...)
+			if !bytes.HasSuffix(existingBytes, []byte("\n")) {
+				combined = append(combined, '\n')
+			}
+			combined = append(combined, '\n')
+			combined = append(combined, data...)
+			ginCtx.Set(apiWebsocketTimelineKey, combined)
+			return
+		}
+	}
+	ginCtx.Set(apiWebsocketTimelineKey, bytes.Clone(data))
+}
+
+func markAPIResponseTimestamp(ginCtx *gin.Context) {
+	if ginCtx == nil {
+		return
+	}
+	if _, exists := ginCtx.Get("API_RESPONSE_TIMESTAMP"); exists {
+		return
+	}
+	ginCtx.Set("API_RESPONSE_TIMESTAMP", time.Now())
+}
+
 func writeHeaders(builder *strings.Builder, headers http.Header) {
 	if builder == nil {
 		return
@@ -285,7 +464,7 @@ func writeHeaders(builder *strings.Builder, headers http.Header) {
 	}
 }
 
-func formatAuthInfo(info upstreamRequestLog) string {
+func formatAuthInfo(info UpstreamRequestLog) string {
 	var parts []string
 	if trimmed := strings.TrimSpace(info.Provider); trimmed != "" {
 		parts = append(parts, fmt.Sprintf("provider=%s", trimmed))
@@ -321,7 +500,7 @@ func formatAuthInfo(info upstreamRequestLog) string {
 	return strings.Join(parts, ", ")
 }
 
-func summarizeErrorBody(contentType string, body []byte) string {
+func SummarizeErrorBody(contentType string, body []byte) string {
 	isHTML := strings.Contains(strings.ToLower(contentType), "text/html")
 	if !isHTML {
 		trimmed := bytes.TrimSpace(bytes.ToLower(body))
@@ -379,7 +558,7 @@ func extractJSONErrorMessage(body []byte) string {
 
 // logWithRequestID returns a logrus Entry with request_id field populated from context.
 // If no request ID is found in context, it returns the standard logger.
-func logWithRequestID(ctx context.Context) *log.Entry {
+func LogWithRequestID(ctx context.Context) *log.Entry {
 	if ctx == nil {
 		return log.NewEntry(log.StandardLogger())
 	}

internal/runtime/executor/helps/payload_helpers.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"encoding/json"
@@ -11,12 +11,12 @@ import (
 	"github.com/tidwall/sjson"
 )
 
-// applyPayloadConfigWithRoot behaves like applyPayloadConfig but treats all parameter
+// ApplyPayloadConfigWithRoot behaves like applyPayloadConfig but treats all parameter
 // paths as relative to the provided root path (for example, "request" for Gemini CLI)
 // and restricts matches to the given protocol when supplied. Defaults are checked
 // against the original payload when provided. requestedModel carries the client-visible
 // model name before alias resolution so payload rules can target aliases precisely.
-func applyPayloadConfigWithRoot(cfg *config.Config, model, protocol, root string, payload, original []byte, requestedModel string) []byte {
+func ApplyPayloadConfigWithRoot(cfg *config.Config, model, protocol, root string, payload, original []byte, requestedModel string) []byte {
 	if cfg == nil || len(payload) == 0 {
 		return payload
 	}
@@ -244,7 +244,7 @@ func payloadRawValue(value any) ([]byte, bool) {
 	}
 }
 
-func payloadRequestedModel(opts cliproxyexecutor.Options, fallback string) string {
+func PayloadRequestedModel(opts cliproxyexecutor.Options, fallback string) string {
 	fallback = strings.TrimSpace(fallback)
 	if len(opts.Metadata) == 0 {
 		return fallback

internal/runtime/executor/helps/proxy_helpers.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"context"
@@ -19,7 +19,7 @@ var (
 	httpClientCacheMutex sync.RWMutex
 )
 
-// newProxyAwareHTTPClient creates an HTTP client with proper proxy configuration priority:
+// NewProxyAwareHTTPClient creates an HTTP client with proper proxy configuration priority:
 // 1. Use auth.ProxyURL if configured (highest priority)
 // 2. Use cfg.ProxyURL if auth proxy is not configured
 // 3. Use RoundTripper from context if neither are configured
@@ -34,7 +34,7 @@ var (
 //
 // Returns:
 //   - *http.Client: An HTTP client with configured proxy or transport
-func newProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
+func NewProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
 	// Priority 1: Use auth.ProxyURL if configured
 	var proxyURL string
 	if auth != nil {
@@ -46,23 +46,18 @@ func newProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *clip
 		proxyURL = strings.TrimSpace(cfg.ProxyURL)
 	}
 
-	// Build cache key from proxy URL (empty string for no proxy)
-	cacheKey := proxyURL
-
-	// Check cache first
-	httpClientCacheMutex.RLock()
-	if cachedClient, ok := httpClientCache[cacheKey]; ok {
-		httpClientCacheMutex.RUnlock()
-		// Return a wrapper with the requested timeout but shared transport
-		if timeout > 0 {
-			return &http.Client{
-				Transport: cachedClient.Transport,
-				Timeout:   timeout,
+	// If we have a proxy URL configured, try cache first to reuse TCP/TLS connections.
+	if proxyURL != "" {
+		httpClientCacheMutex.RLock()
+		if cachedClient, ok := httpClientCache[proxyURL]; ok {
+			httpClientCacheMutex.RUnlock()
+			if timeout > 0 {
+				return &http.Client{Transport: cachedClient.Transport, Timeout: timeout}
 			}
+			return cachedClient
 		}
-		return cachedClient
+		httpClientCacheMutex.RUnlock()
 	}
-	httpClientCacheMutex.RUnlock()
 
 	// Create new client
 	httpClient := &http.Client{}
@@ -77,7 +72,7 @@ func newProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *clip
 			httpClient.Transport = transport
 			// Cache the client
 			httpClientCacheMutex.Lock()
-			httpClientCache[cacheKey] = httpClient
+			httpClientCache[proxyURL] = httpClient
 			httpClientCacheMutex.Unlock()
 			return httpClient
 		}
@@ -90,13 +85,6 @@ func newProxyAwareHTTPClient(ctx context.Context, cfg *config.Config, auth *clip
 		httpClient.Transport = rt
 	}
 
-	// Cache the client for no-proxy case
-	if proxyURL == "" {
-		httpClientCacheMutex.Lock()
-		httpClientCache[cacheKey] = httpClient
-		httpClientCacheMutex.Unlock()
-	}
-
 	return httpClient
 }
 

internal/runtime/executor/helps/proxy_helpers_test.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"context"
@@ -13,7 +13,7 @@ import (
 func TestNewProxyAwareHTTPClientDirectBypassesGlobalProxy(t *testing.T) {
 	t.Parallel()
 
-	client := newProxyAwareHTTPClient(
+	client := NewProxyAwareHTTPClient(
 		context.Background(),
 		&config.Config{SDKConfig: sdkconfig.SDKConfig{ProxyURL: "http://global-proxy.example.com:8080"}},
 		&cliproxyauth.Auth{ProxyURL: "direct"},

internal/runtime/executor/helps/session_id_cache.go

@@ -0,0 +1,92 @@
+package helps
+
+import (
+	"crypto/sha256"
+	"encoding/hex"
+	"sync"
+	"time"
+
+	"github.com/google/uuid"
+)
+
+type sessionIDCacheEntry struct {
+	value  string
+	expire time.Time
+}
+
+var (
+	sessionIDCache            = make(map[string]sessionIDCacheEntry)
+	sessionIDCacheMu          sync.RWMutex
+	sessionIDCacheCleanupOnce sync.Once
+)
+
+const (
+	sessionIDTTL                = time.Hour
+	sessionIDCacheCleanupPeriod = 15 * time.Minute
+)
+
+func startSessionIDCacheCleanup() {
+	go func() {
+		ticker := time.NewTicker(sessionIDCacheCleanupPeriod)
+		defer ticker.Stop()
+		for range ticker.C {
+			purgeExpiredSessionIDs()
+		}
+	}()
+}
+
+func purgeExpiredSessionIDs() {
+	now := time.Now()
+	sessionIDCacheMu.Lock()
+	for key, entry := range sessionIDCache {
+		if !entry.expire.After(now) {
+			delete(sessionIDCache, key)
+		}
+	}
+	sessionIDCacheMu.Unlock()
+}
+
+func sessionIDCacheKey(apiKey string) string {
+	sum := sha256.Sum256([]byte(apiKey))
+	return hex.EncodeToString(sum[:])
+}
+
+// CachedSessionID returns a stable session UUID per apiKey, refreshing the TTL on each access.
+func CachedSessionID(apiKey string) string {
+	if apiKey == "" {
+		return uuid.New().String()
+	}
+
+	sessionIDCacheCleanupOnce.Do(startSessionIDCacheCleanup)
+
+	key := sessionIDCacheKey(apiKey)
+	now := time.Now()
+
+	sessionIDCacheMu.RLock()
+	entry, ok := sessionIDCache[key]
+	valid := ok && entry.value != "" && entry.expire.After(now)
+	sessionIDCacheMu.RUnlock()
+	if valid {
+		sessionIDCacheMu.Lock()
+		entry = sessionIDCache[key]
+		if entry.value != "" && entry.expire.After(now) {
+			entry.expire = now.Add(sessionIDTTL)
+			sessionIDCache[key] = entry
+			sessionIDCacheMu.Unlock()
+			return entry.value
+		}
+		sessionIDCacheMu.Unlock()
+	}
+
+	newID := uuid.New().String()
+
+	sessionIDCacheMu.Lock()
+	entry, ok = sessionIDCache[key]
+	if !ok || entry.value == "" || !entry.expire.After(now) {
+		entry.value = newID
+	}
+	entry.expire = now.Add(sessionIDTTL)
+	sessionIDCache[key] = entry
+	sessionIDCacheMu.Unlock()
+	return entry.value
+}

internal/runtime/executor/helps/thinking_providers.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	_ "github.com/router-for-me/CLIProxyAPI/v6/internal/thinking/provider/antigravity"

internal/runtime/executor/helps/token_helpers.go

@@ -1,9 +1,7 @@
-package executor
+package helps
 
 import (
 	"fmt"
-	"regexp"
-	"strconv"
 	"strings"
 	"sync"
 
@@ -11,100 +9,80 @@ import (
 	"github.com/tiktoken-go/tokenizer"
 )
 
-// tokenizerCache stores tokenizer instances to avoid repeated creation
+// tokenizerCache stores tokenizer instances to avoid repeated creation.
 var tokenizerCache sync.Map
 
-// TokenizerWrapper wraps a tokenizer codec with an adjustment factor for models
-// where tiktoken may not accurately estimate token counts (e.g., Claude models)
-type TokenizerWrapper struct {
-	Codec            tokenizer.Codec
-	AdjustmentFactor float64 // 1.0 means no adjustment, >1.0 means tiktoken underestimates
+type adjustedTokenizer struct {
+	tokenizer.Codec
+	adjustmentFactor float64
 }
 
-// Count returns the token count with adjustment factor applied
-func (tw *TokenizerWrapper) Count(text string) (int, error) {
+func (tw *adjustedTokenizer) Count(text string) (int, error) {
 	count, err := tw.Codec.Count(text)
 	if err != nil {
 		return 0, err
 	}
-	if tw.AdjustmentFactor != 1.0 && tw.AdjustmentFactor > 0 {
-		return int(float64(count) * tw.AdjustmentFactor), nil
+	if tw.adjustmentFactor > 0 && tw.adjustmentFactor != 1.0 {
+		return int(float64(count) * tw.adjustmentFactor), nil
 	}
 	return count, nil
 }
 
-// getTokenizer returns a cached tokenizer for the given model.
-// This improves performance by avoiding repeated tokenizer creation.
-func getTokenizer(model string) (*TokenizerWrapper, error) {
-	// Check cache first
-	if cached, ok := tokenizerCache.Load(model); ok {
-		return cached.(*TokenizerWrapper), nil
+// TokenizerForModel returns a tokenizer codec suitable for an OpenAI-style model id.
+// For Claude-like models, it applies an adjustment factor since tiktoken may underestimate token counts.
+func TokenizerForModel(model string) (tokenizer.Codec, error) {
+	sanitized := strings.ToLower(strings.TrimSpace(model))
+	if cached, ok := tokenizerCache.Load(sanitized); ok {
+		return cached.(tokenizer.Codec), nil
 	}
 
-	// Cache miss, create new tokenizer
-	wrapper, err := tokenizerForModel(model)
+	enc, err := tokenizerForModel(sanitized)
 	if err != nil {
 		return nil, err
 	}
 
-	// Store in cache (use LoadOrStore to handle race conditions)
-	actual, _ := tokenizerCache.LoadOrStore(model, wrapper)
-	return actual.(*TokenizerWrapper), nil
+	actual, _ := tokenizerCache.LoadOrStore(sanitized, enc)
+	return actual.(tokenizer.Codec), nil
 }
 
-// tokenizerForModel returns a tokenizer codec suitable for an OpenAI-style model id.
-// For Claude models, applies a 1.1 adjustment factor since tiktoken may underestimate.
-func tokenizerForModel(model string) (*TokenizerWrapper, error) {
-	sanitized := strings.ToLower(strings.TrimSpace(model))
+func tokenizerForModel(sanitized string) (tokenizer.Codec, error) {
+	if sanitized == "" {
+		return tokenizer.Get(tokenizer.Cl100kBase)
+	}
 
-	// Claude models use cl100k_base with 1.1 adjustment factor
-	// because tiktoken may underestimate Claude's actual token count
+	// Claude models use cl100k_base with an adjustment factor because tiktoken may underestimate.
 	if strings.Contains(sanitized, "claude") || strings.HasPrefix(sanitized, "kiro-") || strings.HasPrefix(sanitized, "amazonq-") {
 		enc, err := tokenizer.Get(tokenizer.Cl100kBase)
 		if err != nil {
 			return nil, err
 		}
-		return &TokenizerWrapper{Codec: enc, AdjustmentFactor: 1.1}, nil
+		return &adjustedTokenizer{Codec: enc, adjustmentFactor: 1.1}, nil
 	}
 
-	var enc tokenizer.Codec
-	var err error
-
 	switch {
-	case sanitized == "":
-		enc, err = tokenizer.Get(tokenizer.Cl100kBase)
-	case strings.HasPrefix(sanitized, "gpt-5.2"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT5)
-	case strings.HasPrefix(sanitized, "gpt-5.1"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT5)
 	case strings.HasPrefix(sanitized, "gpt-5"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT5)
+		return tokenizer.ForModel(tokenizer.GPT5)
 	case strings.HasPrefix(sanitized, "gpt-4.1"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT41)
+		return tokenizer.ForModel(tokenizer.GPT41)
 	case strings.HasPrefix(sanitized, "gpt-4o"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT4o)
+		return tokenizer.ForModel(tokenizer.GPT4o)
 	case strings.HasPrefix(sanitized, "gpt-4"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT4)
+		return tokenizer.ForModel(tokenizer.GPT4)
 	case strings.HasPrefix(sanitized, "gpt-3.5"), strings.HasPrefix(sanitized, "gpt-3"):
-		enc, err = tokenizer.ForModel(tokenizer.GPT35Turbo)
+		return tokenizer.ForModel(tokenizer.GPT35Turbo)
 	case strings.HasPrefix(sanitized, "o1"):
-		enc, err = tokenizer.ForModel(tokenizer.O1)
+		return tokenizer.ForModel(tokenizer.O1)
 	case strings.HasPrefix(sanitized, "o3"):
-		enc, err = tokenizer.ForModel(tokenizer.O3)
+		return tokenizer.ForModel(tokenizer.O3)
 	case strings.HasPrefix(sanitized, "o4"):
-		enc, err = tokenizer.ForModel(tokenizer.O4Mini)
+		return tokenizer.ForModel(tokenizer.O4Mini)
 	default:
-		enc, err = tokenizer.Get(tokenizer.O200kBase)
+		return tokenizer.Get(tokenizer.O200kBase)
 	}
-
-	if err != nil {
-		return nil, err
-	}
-	return &TokenizerWrapper{Codec: enc, AdjustmentFactor: 1.0}, nil
 }
 
-// countOpenAIChatTokens approximates prompt tokens for OpenAI chat completions payloads.
-func countOpenAIChatTokens(enc *TokenizerWrapper, payload []byte) (int64, error) {
+// CountOpenAIChatTokens approximates prompt tokens for OpenAI chat completions payloads.
+func CountOpenAIChatTokens(enc tokenizer.Codec, payload []byte) (int64, error) {
 	if enc == nil {
 		return 0, fmt.Errorf("encoder is nil")
 	}
@@ -128,22 +106,15 @@ func countOpenAIChatTokens(enc *TokenizerWrapper, payload []byte) (int64, error)
 		return 0, nil
 	}
 
-	// Count text tokens
 	count, err := enc.Count(joined)
 	if err != nil {
 		return 0, err
 	}
-
-	// Extract and add image tokens from placeholders
-	imageTokens := extractImageTokens(joined)
-
-	return int64(count) + int64(imageTokens), nil
+	return int64(count), nil
 }
 
-// countClaudeChatTokens approximates prompt tokens for Claude API chat completions payloads.
-// This handles Claude's message format with system, messages, and tools.
-// Image tokens are estimated based on image dimensions when available.
-func countClaudeChatTokens(enc *TokenizerWrapper, payload []byte) (int64, error) {
+// CountClaudeChatTokens approximates prompt tokens for Claude API chat payloads.
+func CountClaudeChatTokens(enc tokenizer.Codec, payload []byte) (int64, error) {
 	if enc == nil {
 		return 0, fmt.Errorf("encoder is nil")
 	}
@@ -153,185 +124,25 @@ func countClaudeChatTokens(enc *TokenizerWrapper, payload []byte) (int64, error)
 
 	root := gjson.ParseBytes(payload)
 	segments := make([]string, 0, 32)
+	imageTokens := 0
 
-	// Collect system prompt (can be string or array of content blocks)
-	collectClaudeSystem(root.Get("system"), &segments)
-
-	// Collect messages
-	collectClaudeMessages(root.Get("messages"), &segments)
-
-	// Collect tools
+	collectClaudeContent(root.Get("system"), &segments, &imageTokens)
+	collectClaudeMessages(root.Get("messages"), &segments, &imageTokens)
 	collectClaudeTools(root.Get("tools"), &segments)
 
 	joined := strings.TrimSpace(strings.Join(segments, "\n"))
 	if joined == "" {
-		return 0, nil
+		return int64(imageTokens), nil
 	}
-
-	// Count text tokens
 	count, err := enc.Count(joined)
 	if err != nil {
 		return 0, err
 	}
-
-	// Extract and add image tokens from placeholders
-	imageTokens := extractImageTokens(joined)
-
-	return int64(count) + int64(imageTokens), nil
+	return int64(count + imageTokens), nil
 }
 
-// imageTokenPattern matches [IMAGE:xxx tokens] format for extracting estimated image tokens
-var imageTokenPattern = regexp.MustCompile(`\[IMAGE:(\d+) tokens\]`)
-
-// extractImageTokens extracts image token estimates from placeholder text.
-// Placeholders are in the format [IMAGE:xxx tokens] where xxx is the estimated token count.
-func extractImageTokens(text string) int {
-	matches := imageTokenPattern.FindAllStringSubmatch(text, -1)
-	total := 0
-	for _, match := range matches {
-		if len(match) > 1 {
-			if tokens, err := strconv.Atoi(match[1]); err == nil {
-				total += tokens
-			}
-		}
-	}
-	return total
-}
-
-// estimateImageTokens calculates estimated tokens for an image based on dimensions.
-// Based on Claude's image token calculation: tokens ≈ (width * height) / 750
-// Minimum 85 tokens, maximum 1590 tokens (for 1568x1568 images).
-func estimateImageTokens(width, height float64) int {
-	if width <= 0 || height <= 0 {
-		// No valid dimensions, use default estimate (medium-sized image)
-		return 1000
-	}
-
-	tokens := int(width * height / 750)
-
-	// Apply bounds
-	if tokens < 85 {
-		tokens = 85
-	}
-	if tokens > 1590 {
-		tokens = 1590
-	}
-
-	return tokens
-}
-
-// collectClaudeSystem extracts text from Claude's system field.
-// System can be a string or an array of content blocks.
-func collectClaudeSystem(system gjson.Result, segments *[]string) {
-	if !system.Exists() {
-		return
-	}
-	if system.Type == gjson.String {
-		addIfNotEmpty(segments, system.String())
-		return
-	}
-	if system.IsArray() {
-		system.ForEach(func(_, block gjson.Result) bool {
-			blockType := block.Get("type").String()
-			if blockType == "text" || blockType == "" {
-				addIfNotEmpty(segments, block.Get("text").String())
-			}
-			// Also handle plain string blocks
-			if block.Type == gjson.String {
-				addIfNotEmpty(segments, block.String())
-			}
-			return true
-		})
-	}
-}
-
-// collectClaudeMessages extracts text from Claude's messages array.
-func collectClaudeMessages(messages gjson.Result, segments *[]string) {
-	if !messages.Exists() || !messages.IsArray() {
-		return
-	}
-	messages.ForEach(func(_, message gjson.Result) bool {
-		addIfNotEmpty(segments, message.Get("role").String())
-		collectClaudeContent(message.Get("content"), segments)
-		return true
-	})
-}
-
-// collectClaudeContent extracts text from Claude's content field.
-// Content can be a string or an array of content blocks.
-// For images, estimates token count based on dimensions when available.
-func collectClaudeContent(content gjson.Result, segments *[]string) {
-	if !content.Exists() {
-		return
-	}
-	if content.Type == gjson.String {
-		addIfNotEmpty(segments, content.String())
-		return
-	}
-	if content.IsArray() {
-		content.ForEach(func(_, part gjson.Result) bool {
-			partType := part.Get("type").String()
-			switch partType {
-			case "text":
-				addIfNotEmpty(segments, part.Get("text").String())
-			case "image":
-				// Estimate image tokens based on dimensions if available
-				source := part.Get("source")
-				if source.Exists() {
-					width := source.Get("width").Float()
-					height := source.Get("height").Float()
-					if width > 0 && height > 0 {
-						tokens := estimateImageTokens(width, height)
-						addIfNotEmpty(segments, fmt.Sprintf("[IMAGE:%d tokens]", tokens))
-					} else {
-						// No dimensions available, use default estimate
-						addIfNotEmpty(segments, "[IMAGE:1000 tokens]")
-					}
-				} else {
-					// No source info, use default estimate
-					addIfNotEmpty(segments, "[IMAGE:1000 tokens]")
-				}
-			case "tool_use":
-				addIfNotEmpty(segments, part.Get("id").String())
-				addIfNotEmpty(segments, part.Get("name").String())
-				if input := part.Get("input"); input.Exists() {
-					addIfNotEmpty(segments, input.Raw)
-				}
-			case "tool_result":
-				addIfNotEmpty(segments, part.Get("tool_use_id").String())
-				collectClaudeContent(part.Get("content"), segments)
-			case "thinking":
-				addIfNotEmpty(segments, part.Get("thinking").String())
-			default:
-				// For unknown types, try to extract any text content
-				if part.Type == gjson.String {
-					addIfNotEmpty(segments, part.String())
-				} else if part.Type == gjson.JSON {
-					addIfNotEmpty(segments, part.Raw)
-				}
-			}
-			return true
-		})
-	}
-}
-
-// collectClaudeTools extracts text from Claude's tools array.
-func collectClaudeTools(tools gjson.Result, segments *[]string) {
-	if !tools.Exists() || !tools.IsArray() {
-		return
-	}
-	tools.ForEach(func(_, tool gjson.Result) bool {
-		addIfNotEmpty(segments, tool.Get("name").String())
-		addIfNotEmpty(segments, tool.Get("description").String())
-		if inputSchema := tool.Get("input_schema"); inputSchema.Exists() {
-			addIfNotEmpty(segments, inputSchema.Raw)
-		}
-		return true
-	})
-}
-
-// buildOpenAIUsageJSON returns a minimal usage structure understood by downstream translators.
-func buildOpenAIUsageJSON(count int64) []byte {
+// BuildOpenAIUsageJSON returns a minimal usage structure understood by downstream translators.
+func BuildOpenAIUsageJSON(count int64) []byte {
 	return []byte(fmt.Sprintf(`{"usage":{"prompt_tokens":%d,"completion_tokens":0,"total_tokens":%d}}`, count, count))
 }
 
@@ -390,6 +201,10 @@ func collectOpenAIContent(content gjson.Result, segments *[]string) {
 	}
 }
 
+func CollectOpenAIContent(content gjson.Result, segments *[]string) {
+	collectOpenAIContent(content, segments)
+}
+
 func collectOpenAIToolCalls(calls gjson.Result, segments *[]string) {
 	if !calls.Exists() || !calls.IsArray() {
 		return
@@ -487,6 +302,98 @@ func appendToolPayload(tool gjson.Result, segments *[]string) {
 	}
 }
 
+func collectClaudeMessages(messages gjson.Result, segments *[]string, imageTokens *int) {
+	if !messages.Exists() || !messages.IsArray() {
+		return
+	}
+	messages.ForEach(func(_, message gjson.Result) bool {
+		addIfNotEmpty(segments, message.Get("role").String())
+		collectClaudeContent(message.Get("content"), segments, imageTokens)
+		return true
+	})
+}
+
+func collectClaudeContent(content gjson.Result, segments *[]string, imageTokens *int) {
+	if !content.Exists() {
+		return
+	}
+	if content.Type == gjson.String {
+		addIfNotEmpty(segments, content.String())
+		return
+	}
+	if content.IsArray() {
+		content.ForEach(func(_, part gjson.Result) bool {
+			partType := part.Get("type").String()
+			switch partType {
+			case "text":
+				addIfNotEmpty(segments, part.Get("text").String())
+			case "image":
+				source := part.Get("source")
+				width := source.Get("width").Float()
+				height := source.Get("height").Float()
+				if imageTokens != nil {
+					*imageTokens += estimateImageTokens(width, height)
+				}
+			case "tool_use":
+				addIfNotEmpty(segments, part.Get("id").String())
+				addIfNotEmpty(segments, part.Get("name").String())
+				if input := part.Get("input"); input.Exists() {
+					addIfNotEmpty(segments, input.Raw)
+				}
+			case "tool_result":
+				addIfNotEmpty(segments, part.Get("tool_use_id").String())
+				collectClaudeContent(part.Get("content"), segments, imageTokens)
+			case "thinking":
+				addIfNotEmpty(segments, part.Get("thinking").String())
+			default:
+				if part.Type == gjson.String {
+					addIfNotEmpty(segments, part.String())
+				} else if part.Type == gjson.JSON {
+					addIfNotEmpty(segments, part.Raw)
+				}
+			}
+			return true
+		})
+		return
+	}
+	if content.Type == gjson.JSON {
+		addIfNotEmpty(segments, content.Raw)
+	}
+}
+
+func collectClaudeTools(tools gjson.Result, segments *[]string) {
+	if !tools.Exists() || !tools.IsArray() {
+		return
+	}
+	tools.ForEach(func(_, tool gjson.Result) bool {
+		addIfNotEmpty(segments, tool.Get("name").String())
+		addIfNotEmpty(segments, tool.Get("description").String())
+		if inputSchema := tool.Get("input_schema"); inputSchema.Exists() {
+			addIfNotEmpty(segments, inputSchema.Raw)
+		}
+		return true
+	})
+}
+
+// estimateImageTokens calculates estimated tokens for an image based on dimensions.
+// Based on Claude's image token calculation: tokens ≈ (width * height) / 750
+// Minimum 85 tokens, maximum 1590 tokens (for 1568x1568 images).
+func estimateImageTokens(width, height float64) int {
+	if width <= 0 || height <= 0 {
+		// No valid dimensions, use default estimate (medium-sized image).
+		return 1000
+	}
+
+	tokens := int(width * height / 750)
+	if tokens < 85 {
+		return 85
+	}
+	if tokens > 1590 {
+		return 1590
+	}
+	return tokens
+}
+
 func addIfNotEmpty(segments *[]string, value string) {
 	if segments == nil {
 		return

internal/runtime/executor/helps/usage_helpers.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"bytes"
@@ -15,7 +15,7 @@ import (
 	"github.com/tidwall/sjson"
 )
 
-type usageReporter struct {
+type UsageReporter struct {
 	provider    string
 	model       string
 	authID      string
@@ -26,9 +26,9 @@ type usageReporter struct {
 	once        sync.Once
 }
 
-func newUsageReporter(ctx context.Context, provider, model string, auth *cliproxyauth.Auth) *usageReporter {
-	apiKey := apiKeyFromContext(ctx)
-	reporter := &usageReporter{
+func NewUsageReporter(ctx context.Context, provider, model string, auth *cliproxyauth.Auth) *UsageReporter {
+	apiKey := APIKeyFromContext(ctx)
+	reporter := &UsageReporter{
 		provider:    provider,
 		model:       model,
 		requestedAt: time.Now(),
@@ -42,24 +42,24 @@ func newUsageReporter(ctx context.Context, provider, model string, auth *cliprox
 	return reporter
 }
 
-func (r *usageReporter) publish(ctx context.Context, detail usage.Detail) {
+func (r *UsageReporter) Publish(ctx context.Context, detail usage.Detail) {
 	r.publishWithOutcome(ctx, detail, false)
 }
 
-func (r *usageReporter) publishFailure(ctx context.Context) {
+func (r *UsageReporter) PublishFailure(ctx context.Context) {
 	r.publishWithOutcome(ctx, usage.Detail{}, true)
 }
 
-func (r *usageReporter) trackFailure(ctx context.Context, errPtr *error) {
+func (r *UsageReporter) TrackFailure(ctx context.Context, errPtr *error) {
 	if r == nil || errPtr == nil {
 		return
 	}
 	if *errPtr != nil {
-		r.publishFailure(ctx)
+		r.PublishFailure(ctx)
 	}
 }
 
-func (r *usageReporter) publishWithOutcome(ctx context.Context, detail usage.Detail, failed bool) {
+func (r *UsageReporter) publishWithOutcome(ctx context.Context, detail usage.Detail, failed bool) {
 	if r == nil {
 		return
 	}
@@ -81,7 +81,7 @@ func (r *usageReporter) publishWithOutcome(ctx context.Context, detail usage.Det
 // It is safe to call multiple times; only the first call wins due to once.Do.
 // This is used to ensure request counting even when upstream responses do not
 // include any usage fields (tokens), especially for streaming paths.
-func (r *usageReporter) ensurePublished(ctx context.Context) {
+func (r *UsageReporter) EnsurePublished(ctx context.Context) {
 	if r == nil {
 		return
 	}
@@ -90,7 +90,7 @@ func (r *usageReporter) ensurePublished(ctx context.Context) {
 	})
 }
 
-func (r *usageReporter) buildRecord(detail usage.Detail, failed bool) usage.Record {
+func (r *UsageReporter) buildRecord(detail usage.Detail, failed bool) usage.Record {
 	if r == nil {
 		return usage.Record{Detail: detail, Failed: failed}
 	}
@@ -108,7 +108,7 @@ func (r *usageReporter) buildRecord(detail usage.Detail, failed bool) usage.Reco
 	}
 }
 
-func (r *usageReporter) latency() time.Duration {
+func (r *UsageReporter) latency() time.Duration {
 	if r == nil || r.requestedAt.IsZero() {
 		return 0
 	}
@@ -119,7 +119,7 @@ func (r *usageReporter) latency() time.Duration {
 	return latency
 }
 
-func apiKeyFromContext(ctx context.Context) string {
+func APIKeyFromContext(ctx context.Context) string {
 	if ctx == nil {
 		return ""
 	}
@@ -184,7 +184,7 @@ func resolveUsageSource(auth *cliproxyauth.Auth, ctxAPIKey string) string {
 	return ""
 }
 
-func parseCodexUsage(data []byte) (usage.Detail, bool) {
+func ParseCodexUsage(data []byte) (usage.Detail, bool) {
 	usageNode := gjson.ParseBytes(data).Get("response.usage")
 	if !usageNode.Exists() {
 		return usage.Detail{}, false
@@ -203,7 +203,7 @@ func parseCodexUsage(data []byte) (usage.Detail, bool) {
 	return detail, true
 }
 
-func parseOpenAIUsage(data []byte) usage.Detail {
+func ParseOpenAIUsage(data []byte) usage.Detail {
 	usageNode := gjson.ParseBytes(data).Get("usage")
 	if !usageNode.Exists() {
 		return usage.Detail{}
@@ -238,7 +238,7 @@ func parseOpenAIUsage(data []byte) usage.Detail {
 	return detail
 }
 
-func parseOpenAIStreamUsage(line []byte) (usage.Detail, bool) {
+func ParseOpenAIStreamUsage(line []byte) (usage.Detail, bool) {
 	payload := jsonPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return usage.Detail{}, false
@@ -247,59 +247,40 @@ func parseOpenAIStreamUsage(line []byte) (usage.Detail, bool) {
 	if !usageNode.Exists() {
 		return usage.Detail{}, false
 	}
+
+	inputNode := usageNode.Get("prompt_tokens")
+	if !inputNode.Exists() {
+		inputNode = usageNode.Get("input_tokens")
+	}
+	outputNode := usageNode.Get("completion_tokens")
+	if !outputNode.Exists() {
+		outputNode = usageNode.Get("output_tokens")
+	}
 	detail := usage.Detail{
-		InputTokens:  usageNode.Get("prompt_tokens").Int(),
-		OutputTokens: usageNode.Get("completion_tokens").Int(),
+		InputTokens:  inputNode.Int(),
+		OutputTokens: outputNode.Int(),
 		TotalTokens:  usageNode.Get("total_tokens").Int(),
 	}
-	if cached := usageNode.Get("prompt_tokens_details.cached_tokens"); cached.Exists() {
+
+	cached := usageNode.Get("prompt_tokens_details.cached_tokens")
+	if !cached.Exists() {
+		cached = usageNode.Get("input_tokens_details.cached_tokens")
+	}
+	if cached.Exists() {
 		detail.CachedTokens = cached.Int()
 	}
-	if reasoning := usageNode.Get("completion_tokens_details.reasoning_tokens"); reasoning.Exists() {
+
+	reasoning := usageNode.Get("completion_tokens_details.reasoning_tokens")
+	if !reasoning.Exists() {
+		reasoning = usageNode.Get("output_tokens_details.reasoning_tokens")
+	}
+	if reasoning.Exists() {
 		detail.ReasoningTokens = reasoning.Int()
 	}
 	return detail, true
 }
 
-func parseOpenAIResponsesUsageDetail(usageNode gjson.Result) usage.Detail {
-	detail := usage.Detail{
-		InputTokens:  usageNode.Get("input_tokens").Int(),
-		OutputTokens: usageNode.Get("output_tokens").Int(),
-		TotalTokens:  usageNode.Get("total_tokens").Int(),
-	}
-	if detail.TotalTokens == 0 {
-		detail.TotalTokens = detail.InputTokens + detail.OutputTokens
-	}
-	if cached := usageNode.Get("input_tokens_details.cached_tokens"); cached.Exists() {
-		detail.CachedTokens = cached.Int()
-	}
-	if reasoning := usageNode.Get("output_tokens_details.reasoning_tokens"); reasoning.Exists() {
-		detail.ReasoningTokens = reasoning.Int()
-	}
-	return detail
-}
-
-func parseOpenAIResponsesUsage(data []byte) usage.Detail {
-	usageNode := gjson.ParseBytes(data).Get("usage")
-	if !usageNode.Exists() {
-		return usage.Detail{}
-	}
-	return parseOpenAIResponsesUsageDetail(usageNode)
-}
-
-func parseOpenAIResponsesStreamUsage(line []byte) (usage.Detail, bool) {
-	payload := jsonPayload(line)
-	if len(payload) == 0 || !gjson.ValidBytes(payload) {
-		return usage.Detail{}, false
-	}
-	usageNode := gjson.GetBytes(payload, "usage")
-	if !usageNode.Exists() {
-		return usage.Detail{}, false
-	}
-	return parseOpenAIResponsesUsageDetail(usageNode), true
-}
-
-func parseClaudeUsage(data []byte) usage.Detail {
+func ParseClaudeUsage(data []byte) usage.Detail {
 	usageNode := gjson.ParseBytes(data).Get("usage")
 	if !usageNode.Exists() {
 		return usage.Detail{}
@@ -317,7 +298,7 @@ func parseClaudeUsage(data []byte) usage.Detail {
 	return detail
 }
 
-func parseClaudeStreamUsage(line []byte) (usage.Detail, bool) {
+func ParseClaudeStreamUsage(line []byte) (usage.Detail, bool) {
 	payload := jsonPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return usage.Detail{}, false
@@ -352,7 +333,7 @@ func parseGeminiFamilyUsageDetail(node gjson.Result) usage.Detail {
 	return detail
 }
 
-func parseGeminiCLIUsage(data []byte) usage.Detail {
+func ParseGeminiCLIUsage(data []byte) usage.Detail {
 	usageNode := gjson.ParseBytes(data)
 	node := usageNode.Get("response.usageMetadata")
 	if !node.Exists() {
@@ -364,7 +345,7 @@ func parseGeminiCLIUsage(data []byte) usage.Detail {
 	return parseGeminiFamilyUsageDetail(node)
 }
 
-func parseGeminiUsage(data []byte) usage.Detail {
+func ParseGeminiUsage(data []byte) usage.Detail {
 	usageNode := gjson.ParseBytes(data)
 	node := usageNode.Get("usageMetadata")
 	if !node.Exists() {
@@ -376,7 +357,7 @@ func parseGeminiUsage(data []byte) usage.Detail {
 	return parseGeminiFamilyUsageDetail(node)
 }
 
-func parseGeminiStreamUsage(line []byte) (usage.Detail, bool) {
+func ParseGeminiStreamUsage(line []byte) (usage.Detail, bool) {
 	payload := jsonPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return usage.Detail{}, false
@@ -391,7 +372,7 @@ func parseGeminiStreamUsage(line []byte) (usage.Detail, bool) {
 	return parseGeminiFamilyUsageDetail(node), true
 }
 
-func parseGeminiCLIStreamUsage(line []byte) (usage.Detail, bool) {
+func ParseGeminiCLIStreamUsage(line []byte) (usage.Detail, bool) {
 	payload := jsonPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return usage.Detail{}, false
@@ -406,7 +387,7 @@ func parseGeminiCLIStreamUsage(line []byte) (usage.Detail, bool) {
 	return parseGeminiFamilyUsageDetail(node), true
 }
 
-func parseAntigravityUsage(data []byte) usage.Detail {
+func ParseAntigravityUsage(data []byte) usage.Detail {
 	usageNode := gjson.ParseBytes(data)
 	node := usageNode.Get("response.usageMetadata")
 	if !node.Exists() {
@@ -421,7 +402,7 @@ func parseAntigravityUsage(data []byte) usage.Detail {
 	return parseGeminiFamilyUsageDetail(node)
 }
 
-func parseAntigravityStreamUsage(line []byte) (usage.Detail, bool) {
+func ParseAntigravityStreamUsage(line []byte) (usage.Detail, bool) {
 	payload := jsonPayload(line)
 	if len(payload) == 0 || !gjson.ValidBytes(payload) {
 		return usage.Detail{}, false
@@ -590,6 +571,10 @@ func isStopChunkWithoutUsage(jsonBytes []byte) bool {
 	return !hasUsageMetadata(jsonBytes)
 }
 
+func JSONPayload(line []byte) []byte {
+	return jsonPayload(line)
+}
+
 func jsonPayload(line []byte) []byte {
 	trimmed := bytes.TrimSpace(line)
 	if len(trimmed) == 0 {

internal/runtime/executor/helps/usage_helpers_test.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"testing"
@@ -9,7 +9,7 @@ import (
 
 func TestParseOpenAIUsageChatCompletions(t *testing.T) {
 	data := []byte(`{"usage":{"prompt_tokens":1,"completion_tokens":2,"total_tokens":3,"prompt_tokens_details":{"cached_tokens":4},"completion_tokens_details":{"reasoning_tokens":5}}}`)
-	detail := parseOpenAIUsage(data)
+	detail := ParseOpenAIUsage(data)
 	if detail.InputTokens != 1 {
 		t.Fatalf("input tokens = %d, want %d", detail.InputTokens, 1)
 	}
@@ -29,7 +29,7 @@ func TestParseOpenAIUsageChatCompletions(t *testing.T) {
 
 func TestParseOpenAIUsageResponses(t *testing.T) {
 	data := []byte(`{"usage":{"input_tokens":10,"output_tokens":20,"total_tokens":30,"input_tokens_details":{"cached_tokens":7},"output_tokens_details":{"reasoning_tokens":9}}}`)
-	detail := parseOpenAIUsage(data)
+	detail := ParseOpenAIUsage(data)
 	if detail.InputTokens != 10 {
 		t.Fatalf("input tokens = %d, want %d", detail.InputTokens, 10)
 	}
@@ -48,7 +48,7 @@ func TestParseOpenAIUsageResponses(t *testing.T) {
 }
 
 func TestUsageReporterBuildRecordIncludesLatency(t *testing.T) {
-	reporter := &usageReporter{
+	reporter := &UsageReporter{
 		provider:    "openai",
 		model:       "gpt-5.4",
 		requestedAt: time.Now().Add(-1500 * time.Millisecond),

internal/runtime/executor/helps/user_id_cache.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"crypto/sha256"
@@ -49,7 +49,7 @@ func userIDCacheKey(apiKey string) string {
 	return hex.EncodeToString(sum[:])
 }
 
-func cachedUserID(apiKey string) string {
+func CachedUserID(apiKey string) string {
 	if apiKey == "" {
 		return generateFakeUserID()
 	}

internal/runtime/executor/helps/user_id_cache_test.go

@@ -1,4 +1,4 @@
-package executor
+package helps
 
 import (
 	"testing"
@@ -14,8 +14,8 @@ func resetUserIDCache() {
 func TestCachedUserID_ReusesWithinTTL(t *testing.T) {
 	resetUserIDCache()
 
-	first := cachedUserID("api-key-1")
-	second := cachedUserID("api-key-1")
+	first := CachedUserID("api-key-1")
+	second := CachedUserID("api-key-1")
 
 	if first == "" {
 		t.Fatal("expected generated user_id to be non-empty")
@@ -28,7 +28,7 @@ func TestCachedUserID_ReusesWithinTTL(t *testing.T) {
 func TestCachedUserID_ExpiresAfterTTL(t *testing.T) {
 	resetUserIDCache()
 
-	expiredID := cachedUserID("api-key-expired")
+	expiredID := CachedUserID("api-key-expired")
 	cacheKey := userIDCacheKey("api-key-expired")
 	userIDCacheMu.Lock()
 	userIDCache[cacheKey] = userIDCacheEntry{
@@ -37,7 +37,7 @@ func TestCachedUserID_ExpiresAfterTTL(t *testing.T) {
 	}
 	userIDCacheMu.Unlock()
 
-	newID := cachedUserID("api-key-expired")
+	newID := CachedUserID("api-key-expired")
 	if newID == expiredID {
 		t.Fatalf("expected expired user_id to be replaced, got %q", newID)
 	}
@@ -49,8 +49,8 @@ func TestCachedUserID_ExpiresAfterTTL(t *testing.T) {
 func TestCachedUserID_IsScopedByAPIKey(t *testing.T) {
 	resetUserIDCache()
 
-	first := cachedUserID("api-key-1")
-	second := cachedUserID("api-key-2")
+	first := CachedUserID("api-key-1")
+	second := CachedUserID("api-key-2")
 
 	if first == second {
 		t.Fatalf("expected different API keys to have different user_ids, got %q", first)
@@ -61,7 +61,7 @@ func TestCachedUserID_RenewsTTLOnHit(t *testing.T) {
 	resetUserIDCache()
 
 	key := "api-key-renew"
-	id := cachedUserID(key)
+	id := CachedUserID(key)
 	cacheKey := userIDCacheKey(key)
 
 	soon := time.Now()
@@ -72,7 +72,7 @@ func TestCachedUserID_RenewsTTLOnHit(t *testing.T) {
 	}
 	userIDCacheMu.Unlock()
 
-	if refreshed := cachedUserID(key); refreshed != id {
+	if refreshed := CachedUserID(key); refreshed != id {
 		t.Fatalf("expected cached user_id to be reused before expiry, got %q", refreshed)
 	}
 

internal/runtime/executor/helps/utls_client.go

@@ -0,0 +1,188 @@
+package helps
+
+import (
+	"net"
+	"net/http"
+	"strings"
+	"sync"
+	"time"
+
+	tls "github.com/refraction-networking/utls"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
+	"github.com/router-for-me/CLIProxyAPI/v6/sdk/proxyutil"
+	log "github.com/sirupsen/logrus"
+	"golang.org/x/net/http2"
+	"golang.org/x/net/proxy"
+)
+
+// utlsRoundTripper implements http.RoundTripper using utls with Chrome fingerprint
+// to bypass Cloudflare's TLS fingerprinting on Anthropic domains.
+type utlsRoundTripper struct {
+	mu          sync.Mutex
+	connections map[string]*http2.ClientConn
+	pending     map[string]*sync.Cond
+	dialer      proxy.Dialer
+}
+
+func newUtlsRoundTripper(proxyURL string) *utlsRoundTripper {
+	var dialer proxy.Dialer = proxy.Direct
+	if proxyURL != "" {
+		proxyDialer, mode, errBuild := proxyutil.BuildDialer(proxyURL)
+		if errBuild != nil {
+			log.Errorf("utls: failed to configure proxy dialer for %q: %v", proxyURL, errBuild)
+		} else if mode != proxyutil.ModeInherit && proxyDialer != nil {
+			dialer = proxyDialer
+		}
+	}
+	return &utlsRoundTripper{
+		connections: make(map[string]*http2.ClientConn),
+		pending:     make(map[string]*sync.Cond),
+		dialer:      dialer,
+	}
+}
+
+func (t *utlsRoundTripper) getOrCreateConnection(host, addr string) (*http2.ClientConn, error) {
+	t.mu.Lock()
+
+	if h2Conn, ok := t.connections[host]; ok && h2Conn.CanTakeNewRequest() {
+		t.mu.Unlock()
+		return h2Conn, nil
+	}
+
+	if cond, ok := t.pending[host]; ok {
+		cond.Wait()
+		if h2Conn, ok := t.connections[host]; ok && h2Conn.CanTakeNewRequest() {
+			t.mu.Unlock()
+			return h2Conn, nil
+		}
+	}
+
+	cond := sync.NewCond(&t.mu)
+	t.pending[host] = cond
+	t.mu.Unlock()
+
+	h2Conn, err := t.createConnection(host, addr)
+
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	delete(t.pending, host)
+	cond.Broadcast()
+
+	if err != nil {
+		return nil, err
+	}
+
+	t.connections[host] = h2Conn
+	return h2Conn, nil
+}
+
+func (t *utlsRoundTripper) createConnection(host, addr string) (*http2.ClientConn, error) {
+	conn, err := t.dialer.Dial("tcp", addr)
+	if err != nil {
+		return nil, err
+	}
+
+	tlsConfig := &tls.Config{ServerName: host}
+	tlsConn := tls.UClient(conn, tlsConfig, tls.HelloChrome_Auto)
+
+	if err := tlsConn.Handshake(); err != nil {
+		conn.Close()
+		return nil, err
+	}
+
+	tr := &http2.Transport{}
+	h2Conn, err := tr.NewClientConn(tlsConn)
+	if err != nil {
+		tlsConn.Close()
+		return nil, err
+	}
+
+	return h2Conn, nil
+}
+
+func (t *utlsRoundTripper) RoundTrip(req *http.Request) (*http.Response, error) {
+	hostname := req.URL.Hostname()
+	port := req.URL.Port()
+	if port == "" {
+		port = "443"
+	}
+	addr := net.JoinHostPort(hostname, port)
+
+	h2Conn, err := t.getOrCreateConnection(hostname, addr)
+	if err != nil {
+		return nil, err
+	}
+
+	resp, err := h2Conn.RoundTrip(req)
+	if err != nil {
+		t.mu.Lock()
+		if cached, ok := t.connections[hostname]; ok && cached == h2Conn {
+			delete(t.connections, hostname)
+		}
+		t.mu.Unlock()
+		return nil, err
+	}
+
+	return resp, nil
+}
+
+// anthropicHosts contains the hosts that should use utls Chrome TLS fingerprint.
+var anthropicHosts = map[string]struct{}{
+	"api.anthropic.com": {},
+}
+
+// fallbackRoundTripper uses utls for Anthropic HTTPS hosts and falls back to
+// standard transport for all other requests (non-HTTPS or non-Anthropic hosts).
+type fallbackRoundTripper struct {
+	utls     *utlsRoundTripper
+	fallback http.RoundTripper
+}
+
+func (f *fallbackRoundTripper) RoundTrip(req *http.Request) (*http.Response, error) {
+	if req.URL.Scheme == "https" {
+		if _, ok := anthropicHosts[strings.ToLower(req.URL.Hostname())]; ok {
+			return f.utls.RoundTrip(req)
+		}
+	}
+	return f.fallback.RoundTrip(req)
+}
+
+// NewUtlsHTTPClient creates an HTTP client using utls Chrome TLS fingerprint.
+// Use this for Claude API requests to match real Claude Code's TLS behavior.
+// Falls back to standard transport for non-HTTPS requests.
+func NewUtlsHTTPClient(cfg *config.Config, auth *cliproxyauth.Auth, timeout time.Duration) *http.Client {
+	var proxyURL string
+	if auth != nil {
+		proxyURL = strings.TrimSpace(auth.ProxyURL)
+	}
+	if proxyURL == "" && cfg != nil {
+		proxyURL = strings.TrimSpace(cfg.ProxyURL)
+	}
+
+	utlsRT := newUtlsRoundTripper(proxyURL)
+
+	var standardTransport http.RoundTripper = &http.Transport{
+		DialContext: (&net.Dialer{
+			Timeout:   30 * time.Second,
+			KeepAlive: 30 * time.Second,
+		}).DialContext,
+	}
+	if proxyURL != "" {
+		if transport := buildProxyTransport(proxyURL); transport != nil {
+			standardTransport = transport
+		}
+	}
+
+	client := &http.Client{
+		Transport: &fallbackRoundTripper{
+			utls:     utlsRT,
+			fallback: standardTransport,
+		},
+	}
+	if timeout > 0 {
+		client.Timeout = timeout
+	}
+	return client
+}

internal/runtime/executor/iflow_executor.go

@@ -16,6 +16,7 @@ import (
 	"github.com/google/uuid"
 	iflowauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/iflow"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -66,7 +67,7 @@ func (e *IFlowExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Auth
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -86,8 +87,8 @@ func (e *IFlowExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		baseURL = iflowauth.DefaultAPIBaseURL
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("openai")
@@ -106,8 +107,8 @@ func (e *IFlowExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 	}
 
 	body = preserveReasoningContentInMessages(body)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 
 	endpoint := strings.TrimSuffix(baseURL, "/") + iflowDefaultEndpoint
 
@@ -116,13 +117,18 @@ func (e *IFlowExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		return resp, err
 	}
 	applyIFlowHeaders(httpReq, apiKey, false)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
 		authID = auth.ID
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       endpoint,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -134,10 +140,10 @@ func (e *IFlowExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -145,25 +151,25 @@ func (e *IFlowExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, re
 			log.Errorf("iflow executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseOpenAIUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseOpenAIUsage(data))
 	// Ensure usage is recorded even if upstream omits usage metadata.
-	reporter.ensurePublished(ctx)
+	reporter.EnsurePublished(ctx)
 
 	var param any
 	// Note: TranslateNonStream uses req.Model (original with suffix) to preserve
@@ -189,8 +195,8 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		baseURL = iflowauth.DefaultAPIBaseURL
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("openai")
@@ -214,8 +220,8 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 	if toolsResult.Exists() && toolsResult.IsArray() && len(toolsResult.Array()) == 0 {
 		body = ensureToolsArray(body)
 	}
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 
 	endpoint := strings.TrimSuffix(baseURL, "/") + iflowDefaultEndpoint
 
@@ -224,13 +230,18 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		return nil, err
 	}
 	applyIFlowHeaders(httpReq, apiKey, true)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
 		authID = auth.ID
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       endpoint,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -242,21 +253,21 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
 
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		data, _ := io.ReadAll(httpResp.Body)
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("iflow executor: close response body error: %v", errClose)
 		}
-		appendAPIResponseChunk(ctx, e.cfg, data)
-		logWithRequestID(ctx).Debugf("request error, error status: %d error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), data))
 		err = statusErr{code: httpResp.StatusCode, msg: string(data)}
 		return nil, err
 	}
@@ -275,9 +286,9 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseOpenAIStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseOpenAIStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(line), &param)
 			for i := range chunks {
@@ -285,12 +296,12 @@ func (e *IFlowExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Au
 			}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 		// Guarantee a usage record exists even if the stream never emitted usage data.
-		reporter.ensurePublished(ctx)
+		reporter.EnsurePublished(ctx)
 	}()
 
 	return &cliproxyexecutor.StreamResult{Headers: httpResp.Header.Clone(), Chunks: out}, nil
@@ -303,17 +314,17 @@ func (e *IFlowExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Auth
 	to := sdktranslator.FromString("openai")
 	body := sdktranslator.TranslateRequest(from, to, baseModel, req.Payload, false)
 
-	enc, err := tokenizerForModel(baseModel)
+	enc, err := helps.TokenizerForModel(baseModel)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("iflow executor: tokenizer init failed: %w", err)
 	}
 
-	count, err := countOpenAIChatTokens(enc, body)
+	count, err := helps.CountOpenAIChatTokens(enc, body)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("iflow executor: token counting failed: %w", err)
 	}
 
-	usageJSON := buildOpenAIUsageJSON(count)
+	usageJSON := helps.BuildOpenAIUsageJSON(count)
 	translated := sdktranslator.TranslateTokenCount(ctx, to, from, count, usageJSON)
 	return cliproxyexecutor.Response{Payload: translated}, nil
 }

internal/runtime/executor/kimi_executor.go

@@ -15,7 +15,9 @@ import (
 
 	kimiauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/kimi"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
@@ -45,6 +47,11 @@ func (e *KimiExecutor) PrepareRequest(req *http.Request, auth *cliproxyauth.Auth
 	if strings.TrimSpace(token) != "" {
 		req.Header.Set("Authorization", "Bearer "+token)
 	}
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(req, attrs)
 	return nil
 }
 
@@ -60,7 +67,7 @@ func (e *KimiExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Auth,
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -76,8 +83,8 @@ func (e *KimiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 
 	token := kimiCreds(auth)
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	to := sdktranslator.FromString("openai")
 	originalPayloadSource := req.Payload
@@ -100,8 +107,8 @@ func (e *KimiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, err = normalizeKimiToolMessageLinks(body)
 	if err != nil {
 		return resp, err
@@ -113,13 +120,18 @@ func (e *KimiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		return resp, err
 	}
 	applyKimiHeadersWithAuth(httpReq, token, false, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
 		authID = auth.ID
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -131,10 +143,10 @@ func (e *KimiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -142,21 +154,21 @@ func (e *KimiExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 			log.Errorf("kimi executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseOpenAIUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseOpenAIUsage(data))
 	var param any
 	// Note: TranslateNonStream uses req.Model (original with suffix) to preserve
 	// the original model name in the response for client compatibility.
@@ -176,8 +188,8 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 	token := kimiCreds(auth)
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	to := sdktranslator.FromString("openai")
 	originalPayloadSource := req.Payload
@@ -204,8 +216,8 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 	if err != nil {
 		return nil, fmt.Errorf("kimi executor: failed to set stream_options in payload: %w", err)
 	}
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
 	body, err = normalizeKimiToolMessageLinks(body)
 	if err != nil {
 		return nil, err
@@ -217,13 +229,18 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		return nil, err
 	}
 	applyKimiHeadersWithAuth(httpReq, token, true, auth)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authID, authLabel, authType, authValue string
 	if auth != nil {
 		authID = auth.ID
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -235,17 +252,17 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("kimi executor: close response body error: %v", errClose)
 		}
@@ -265,9 +282,9 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseOpenAIStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseOpenAIStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(line), &param)
 			for i := range chunks {
@@ -279,8 +296,8 @@ func (e *KimiExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 			out <- cliproxyexecutor.StreamChunk{Payload: doneChunks[i]}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()

internal/runtime/executor/openai_compat_executor.go

@@ -11,6 +11,7 @@ import (
 	"time"
 
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
@@ -65,15 +66,15 @@ func (e *OpenAICompatExecutor) HttpRequest(ctx context.Context, auth *cliproxyau
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
 func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options) (resp cliproxyexecutor.Response, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	baseURL, apiKey := e.resolveCredentials(auth)
 	if baseURL == "" {
@@ -95,8 +96,8 @@ func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.A
 	originalPayload := originalPayloadSource
 	originalTranslated := sdktranslator.TranslateRequest(from, to, baseModel, originalPayload, opts.Stream)
 	translated := sdktranslator.TranslateRequest(from, to, baseModel, req.Payload, opts.Stream)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	translated = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", translated, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	translated = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", translated, originalTranslated, requestedModel)
 	if opts.Alt == "responses/compact" {
 		if updated, errDelete := sjson.DeleteBytes(translated, "stream"); errDelete == nil {
 			translated = updated
@@ -129,7 +130,7 @@ func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.A
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -141,10 +142,10 @@ func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.A
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -152,23 +153,23 @@ func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.A
 			log.Errorf("openai compat executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: httpResp.StatusCode, msg: string(b)}
 		return resp, err
 	}
 	body, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, body)
-	reporter.publish(ctx, parseOpenAIUsage(body))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, body)
+	reporter.Publish(ctx, helps.ParseOpenAIUsage(body))
 	// Ensure we at least record the request even if upstream doesn't return usage
-	reporter.ensurePublished(ctx)
+	reporter.EnsurePublished(ctx)
 	// Translate response back to source format when needed
 	var param any
 	out := sdktranslator.TranslateNonStream(ctx, to, from, req.Model, opts.OriginalRequest, translated, body, &param)
@@ -179,8 +180,8 @@ func (e *OpenAICompatExecutor) Execute(ctx context.Context, auth *cliproxyauth.A
 func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Auth, req cliproxyexecutor.Request, opts cliproxyexecutor.Options) (_ *cliproxyexecutor.StreamResult, err error) {
 	baseModel := thinking.ParseSuffix(req.Model).ModelName
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	baseURL, apiKey := e.resolveCredentials(auth)
 	if baseURL == "" {
@@ -197,8 +198,8 @@ func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxy
 	originalPayload := originalPayloadSource
 	originalTranslated := sdktranslator.TranslateRequest(from, to, baseModel, originalPayload, true)
 	translated := sdktranslator.TranslateRequest(from, to, baseModel, req.Payload, true)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	translated = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", translated, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	translated = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", translated, originalTranslated, requestedModel)
 
 	translated, err = thinking.ApplyThinking(translated, req.Model, from.String(), to.String(), e.Identifier())
 	if err != nil {
@@ -232,7 +233,7 @@ func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxy
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -244,17 +245,17 @@ func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxy
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d, error message: %s", httpResp.StatusCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("openai compat executor: close response body error: %v", errClose)
 		}
@@ -274,9 +275,9 @@ func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxy
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseOpenAIStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseOpenAIStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			if len(line) == 0 {
 				continue
@@ -294,12 +295,12 @@ func (e *OpenAICompatExecutor) ExecuteStream(ctx context.Context, auth *cliproxy
 			}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 		// Ensure we record the request if no usage chunk was ever seen
-		reporter.ensurePublished(ctx)
+		reporter.EnsurePublished(ctx)
 	}()
 	return &cliproxyexecutor.StreamResult{Headers: httpResp.Header.Clone(), Chunks: out}, nil
 }
@@ -318,17 +319,17 @@ func (e *OpenAICompatExecutor) CountTokens(ctx context.Context, auth *cliproxyau
 		return cliproxyexecutor.Response{}, err
 	}
 
-	enc, err := tokenizerForModel(modelForCounting)
+	enc, err := helps.TokenizerForModel(modelForCounting)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("openai compat executor: tokenizer init failed: %w", err)
 	}
 
-	count, err := countOpenAIChatTokens(enc, translated)
+	count, err := helps.CountOpenAIChatTokens(enc, translated)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("openai compat executor: token counting failed: %w", err)
 	}
 
-	usageJSON := buildOpenAIUsageJSON(count)
+	usageJSON := helps.BuildOpenAIUsageJSON(count)
 	translatedUsage := sdktranslator.TranslateTokenCount(ctx, to, from, count, usageJSON)
 	return cliproxyexecutor.Response{Payload: translatedUsage}, nil
 }

internal/runtime/executor/qwen_executor.go

@@ -13,7 +13,9 @@ import (
 
 	qwenauth "github.com/router-for-me/CLIProxyAPI/v6/internal/auth/qwen"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/runtime/executor/helps"
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/util"
 	cliproxyauth "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/auth"
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 	sdktranslator "github.com/router-for-me/CLIProxyAPI/v6/sdk/translator"
@@ -23,11 +25,13 @@ import (
 )
 
 const (
-	qwenUserAgent       = "QwenCode/0.10.3 (darwin; arm64)"
+	qwenUserAgent       = "QwenCode/0.13.2 (darwin; arm64)"
 	qwenRateLimitPerMin = 60          // 60 requests per minute per credential
 	qwenRateLimitWindow = time.Minute // sliding window duration
 )
 
+var qwenDefaultSystemMessage = []byte(`{"role":"system","content":[{"type":"text","text":"","cache_control":{"type":"ephemeral"}}]}`)
+
 // qwenBeijingLoc caches the Beijing timezone to avoid repeated LoadLocation syscalls.
 var qwenBeijingLoc = func() *time.Location {
 	loc, err := time.LoadLocation("Asia/Shanghai")
@@ -154,7 +158,7 @@ func wrapQwenError(ctx context.Context, httpCode int, body []byte) (errCode int,
 		errCode = http.StatusTooManyRequests // Map to 429 to trigger quota logic
 		cooldown := timeUntilNextDay()
 		retryAfter = &cooldown
-		logWithRequestID(ctx).Warnf("qwen quota exceeded (http %d -> %d), cooling down until tomorrow (%v)", httpCode, errCode, cooldown)
+		helps.LogWithRequestID(ctx).Warnf("qwen quota exceeded (http %d -> %d), cooling down until tomorrow (%v)", httpCode, errCode, cooldown)
 	}
 	return errCode, retryAfter
 }
@@ -168,6 +172,36 @@ func timeUntilNextDay() time.Duration {
 	return tomorrow.Sub(now)
 }
 
+// ensureQwenSystemMessage prepends a default system message if none exists in "messages".
+func ensureQwenSystemMessage(payload []byte) ([]byte, error) {
+	messages := gjson.GetBytes(payload, "messages")
+	if messages.Exists() && messages.IsArray() {
+		var buf bytes.Buffer
+		buf.WriteByte('[')
+		buf.Write(qwenDefaultSystemMessage)
+		for _, msg := range messages.Array() {
+			buf.WriteByte(',')
+			buf.WriteString(msg.Raw)
+		}
+		buf.WriteByte(']')
+		updated, errSet := sjson.SetRawBytes(payload, "messages", buf.Bytes())
+		if errSet != nil {
+			return nil, fmt.Errorf("qwen executor: set default system message failed: %w", errSet)
+		}
+		return updated, nil
+	}
+
+	var buf bytes.Buffer
+	buf.WriteByte('[')
+	buf.Write(qwenDefaultSystemMessage)
+	buf.WriteByte(']')
+	updated, errSet := sjson.SetRawBytes(payload, "messages", buf.Bytes())
+	if errSet != nil {
+		return nil, fmt.Errorf("qwen executor: set default system message failed: %w", errSet)
+	}
+	return updated, nil
+}
+
 // QwenExecutor is a stateless executor for Qwen Code using OpenAI-compatible chat completions.
 // If access token is unavailable, it falls back to legacy via ClientAdapter.
 type QwenExecutor struct {
@@ -202,7 +236,7 @@ func (e *QwenExecutor) HttpRequest(ctx context.Context, auth *cliproxyauth.Auth,
 	if err := e.PrepareRequest(httpReq, auth); err != nil {
 		return nil, err
 	}
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	return httpClient.Do(httpReq)
 }
 
@@ -217,7 +251,7 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		authID = auth.ID
 	}
 	if err := checkQwenRateLimit(authID); err != nil {
-		logWithRequestID(ctx).Warnf("qwen rate limit exceeded for credential %s", redactAuthID(authID))
+		helps.LogWithRequestID(ctx).Warnf("qwen rate limit exceeded for credential %s", redactAuthID(authID))
 		return resp, err
 	}
 
@@ -228,8 +262,8 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		baseURL = "https://portal.qwen.ai/v1"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("openai")
@@ -247,8 +281,12 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		return resp, err
 	}
 
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	body, err = ensureQwenSystemMessage(body)
+	if err != nil {
+		return resp, err
+	}
 
 	url := strings.TrimSuffix(baseURL, "/") + "/chat/completions"
 	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(body))
@@ -256,12 +294,17 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		return resp, err
 	}
 	applyQwenHeaders(httpReq, token, false)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authLabel, authType, authValue string
 	if auth != nil {
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -273,10 +316,10 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
 	defer func() {
@@ -284,23 +327,23 @@ func (e *QwenExecutor) Execute(ctx context.Context, auth *cliproxyauth.Auth, req
 			log.Errorf("qwen executor: close response body error: %v", errClose)
 		}
 	}()
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
 
 		errCode, retryAfter := wrapQwenError(ctx, httpResp.StatusCode, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d (mapped: %d), error message: %s", httpResp.StatusCode, errCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d (mapped: %d), error message: %s", httpResp.StatusCode, errCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		err = statusErr{code: errCode, msg: string(b), retryAfter: retryAfter}
 		return resp, err
 	}
 	data, err := io.ReadAll(httpResp.Body)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return resp, err
 	}
-	appendAPIResponseChunk(ctx, e.cfg, data)
-	reporter.publish(ctx, parseOpenAIUsage(data))
+	helps.AppendAPIResponseChunk(ctx, e.cfg, data)
+	reporter.Publish(ctx, helps.ParseOpenAIUsage(data))
 	var param any
 	// Note: TranslateNonStream uses req.Model (original with suffix) to preserve
 	// the original model name in the response for client compatibility.
@@ -320,7 +363,7 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		authID = auth.ID
 	}
 	if err := checkQwenRateLimit(authID); err != nil {
-		logWithRequestID(ctx).Warnf("qwen rate limit exceeded for credential %s", redactAuthID(authID))
+		helps.LogWithRequestID(ctx).Warnf("qwen rate limit exceeded for credential %s", redactAuthID(authID))
 		return nil, err
 	}
 
@@ -331,8 +374,8 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		baseURL = "https://portal.qwen.ai/v1"
 	}
 
-	reporter := newUsageReporter(ctx, e.Identifier(), baseModel, auth)
-	defer reporter.trackFailure(ctx, &err)
+	reporter := helps.NewUsageReporter(ctx, e.Identifier(), baseModel, auth)
+	defer reporter.TrackFailure(ctx, &err)
 
 	from := opts.SourceFormat
 	to := sdktranslator.FromString("openai")
@@ -350,15 +393,19 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		return nil, err
 	}
 
-	toolsResult := gjson.GetBytes(body, "tools")
+	// toolsResult := gjson.GetBytes(body, "tools")
 	// I'm addressing the Qwen3 "poisoning" issue, which is caused by the model needing a tool to be defined. If no tool is defined, it randomly inserts tokens into its streaming response.
 	// This will have no real consequences. It's just to scare Qwen3.
-	if (toolsResult.IsArray() && len(toolsResult.Array()) == 0) || !toolsResult.Exists() {
-		body, _ = sjson.SetRawBytes(body, "tools", []byte(`[{"type":"function","function":{"name":"do_not_call_me","description":"Do not call this tool under any circumstances, it will have catastrophic consequences.","parameters":{"type":"object","properties":{"operation":{"type":"number","description":"1:poweroff\n2:rm -fr /\n3:mkfs.ext4 /dev/sda1"}},"required":["operation"]}}}]`))
-	}
+	// if (toolsResult.IsArray() && len(toolsResult.Array()) == 0) || !toolsResult.Exists() {
+	// 	body, _ = sjson.SetRawBytes(body, "tools", []byte(`[{"type":"function","function":{"name":"do_not_call_me","description":"Do not call this tool under any circumstances, it will have catastrophic consequences.","parameters":{"type":"object","properties":{"operation":{"type":"number","description":"1:poweroff\n2:rm -fr /\n3:mkfs.ext4 /dev/sda1"}},"required":["operation"]}}}]`))
+	// }
 	body, _ = sjson.SetBytes(body, "stream_options.include_usage", true)
-	requestedModel := payloadRequestedModel(opts, req.Model)
-	body = applyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	requestedModel := helps.PayloadRequestedModel(opts, req.Model)
+	body = helps.ApplyPayloadConfigWithRoot(e.cfg, baseModel, to.String(), "", body, originalTranslated, requestedModel)
+	body, err = ensureQwenSystemMessage(body)
+	if err != nil {
+		return nil, err
+	}
 
 	url := strings.TrimSuffix(baseURL, "/") + "/chat/completions"
 	httpReq, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(body))
@@ -366,12 +413,17 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		return nil, err
 	}
 	applyQwenHeaders(httpReq, token, true)
+	var attrs map[string]string
+	if auth != nil {
+		attrs = auth.Attributes
+	}
+	util.ApplyCustomHeadersFromAttrs(httpReq, attrs)
 	var authLabel, authType, authValue string
 	if auth != nil {
 		authLabel = auth.Label
 		authType, authValue = auth.AccountInfo()
 	}
-	recordAPIRequest(ctx, e.cfg, upstreamRequestLog{
+	helps.RecordAPIRequest(ctx, e.cfg, helps.UpstreamRequestLog{
 		URL:       url,
 		Method:    http.MethodPost,
 		Headers:   httpReq.Header.Clone(),
@@ -383,19 +435,19 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		AuthValue: authValue,
 	})
 
-	httpClient := newProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
+	httpClient := helps.NewProxyAwareHTTPClient(ctx, e.cfg, auth, 0)
 	httpResp, err := httpClient.Do(httpReq)
 	if err != nil {
-		recordAPIResponseError(ctx, e.cfg, err)
+		helps.RecordAPIResponseError(ctx, e.cfg, err)
 		return nil, err
 	}
-	recordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
+	helps.RecordAPIResponseMetadata(ctx, e.cfg, httpResp.StatusCode, httpResp.Header.Clone())
 	if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
 		b, _ := io.ReadAll(httpResp.Body)
-		appendAPIResponseChunk(ctx, e.cfg, b)
+		helps.AppendAPIResponseChunk(ctx, e.cfg, b)
 
 		errCode, retryAfter := wrapQwenError(ctx, httpResp.StatusCode, b)
-		logWithRequestID(ctx).Debugf("request error, error status: %d (mapped: %d), error message: %s", httpResp.StatusCode, errCode, summarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
+		helps.LogWithRequestID(ctx).Debugf("request error, error status: %d (mapped: %d), error message: %s", httpResp.StatusCode, errCode, helps.SummarizeErrorBody(httpResp.Header.Get("Content-Type"), b))
 		if errClose := httpResp.Body.Close(); errClose != nil {
 			log.Errorf("qwen executor: close response body error: %v", errClose)
 		}
@@ -415,9 +467,9 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 		var param any
 		for scanner.Scan() {
 			line := scanner.Bytes()
-			appendAPIResponseChunk(ctx, e.cfg, line)
-			if detail, ok := parseOpenAIStreamUsage(line); ok {
-				reporter.publish(ctx, detail)
+			helps.AppendAPIResponseChunk(ctx, e.cfg, line)
+			if detail, ok := helps.ParseOpenAIStreamUsage(line); ok {
+				reporter.Publish(ctx, detail)
 			}
 			chunks := sdktranslator.TranslateStream(ctx, to, from, req.Model, opts.OriginalRequest, body, bytes.Clone(line), &param)
 			for i := range chunks {
@@ -429,8 +481,8 @@ func (e *QwenExecutor) ExecuteStream(ctx context.Context, auth *cliproxyauth.Aut
 			out <- cliproxyexecutor.StreamChunk{Payload: doneChunks[i]}
 		}
 		if errScan := scanner.Err(); errScan != nil {
-			recordAPIResponseError(ctx, e.cfg, errScan)
-			reporter.publishFailure(ctx)
+			helps.RecordAPIResponseError(ctx, e.cfg, errScan)
+			reporter.PublishFailure(ctx)
 			out <- cliproxyexecutor.StreamChunk{Err: errScan}
 		}
 	}()
@@ -449,17 +501,17 @@ func (e *QwenExecutor) CountTokens(ctx context.Context, auth *cliproxyauth.Auth,
 		modelName = baseModel
 	}
 
-	enc, err := tokenizerForModel(modelName)
+	enc, err := helps.TokenizerForModel(modelName)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("qwen executor: tokenizer init failed: %w", err)
 	}
 
-	count, err := countOpenAIChatTokens(enc, body)
+	count, err := helps.CountOpenAIChatTokens(enc, body)
 	if err != nil {
 		return cliproxyexecutor.Response{}, fmt.Errorf("qwen executor: token counting failed: %w", err)
 	}
 
-	usageJSON := buildOpenAIUsageJSON(count)
+	usageJSON := helps.BuildOpenAIUsageJSON(count)
 	translated := sdktranslator.TranslateTokenCount(ctx, to, from, count, usageJSON)
 	return cliproxyexecutor.Response{Payload: translated}, nil
 }
@@ -508,16 +560,15 @@ func applyQwenHeaders(r *http.Request, token string, stream bool) {
 	r.Header.Set("Content-Type", "application/json")
 	r.Header.Set("Authorization", "Bearer "+token)
 	r.Header.Set("User-Agent", qwenUserAgent)
-	r.Header.Set("X-Dashscope-Useragent", qwenUserAgent)
+	r.Header["X-DashScope-UserAgent"] = []string{qwenUserAgent}
 	r.Header.Set("X-Stainless-Runtime-Version", "v22.17.0")
-	r.Header.Set("Sec-Fetch-Mode", "cors")
 	r.Header.Set("X-Stainless-Lang", "js")
 	r.Header.Set("X-Stainless-Arch", "arm64")
 	r.Header.Set("X-Stainless-Package-Version", "5.11.0")
-	r.Header.Set("X-Dashscope-Cachecontrol", "enable")
+	r.Header["X-DashScope-CacheControl"] = []string{"enable"}
 	r.Header.Set("X-Stainless-Retry-Count", "0")
 	r.Header.Set("X-Stainless-Os", "MacOS")
-	r.Header.Set("X-Dashscope-Authtype", "qwen-oauth")
+	r.Header["X-DashScope-AuthType"] = []string{"qwen-oauth"}
 	r.Header.Set("X-Stainless-Runtime", "node")
 
 	if stream {

internal/store/gitstore.go

@@ -446,6 +446,7 @@ func (s *GitTokenStore) readAuthFile(path, baseDir string) (*cliproxyauth.Auth,
 	if email, ok := metadata["email"].(string); ok && email != "" {
 		auth.Attributes["email"] = email
 	}
+	cliproxyauth.ApplyCustomHeadersFromMetadata(auth)
 	return auth, nil
 }
 

internal/store/objectstore.go

@@ -595,6 +595,7 @@ func (s *ObjectTokenStore) readAuthFile(path, baseDir string) (*cliproxyauth.Aut
 		LastRefreshedAt:  time.Time{},
 		NextRefreshAfter: time.Time{},
 	}
+	cliproxyauth.ApplyCustomHeadersFromMetadata(auth)
 	return auth, nil
 }
 

internal/store/postgresstore.go

@@ -310,6 +310,7 @@ func (s *PostgresStore) List(ctx context.Context) ([]*cliproxyauth.Auth, error)
 			LastRefreshedAt:  time.Time{},
 			NextRefreshAfter: time.Time{},
 		}
+		cliproxyauth.ApplyCustomHeadersFromMetadata(auth)
 		auths = append(auths, auth)
 	}
 	if err = rows.Err(); err != nil {

internal/translator/antigravity/claude/antigravity_claude_request.go

@@ -330,32 +330,45 @@ func ConvertClaudeRequestToAntigravity(modelName string, inputRawJSON []byte, _
 					}
 				}
 
-				// Reorder parts for 'model' role to ensure thinking block is first
+				// Reorder parts for 'model' role:
+				// 1. Thinking parts first (Antigravity API requirement)
+				// 2. Regular parts (text, inlineData, etc.)
+				// 3. FunctionCall parts last
+				//
+				// Moving functionCall parts to the end prevents tool_use↔tool_result
+				// pairing breakage: the Antigravity API internally splits model messages
+				// at functionCall boundaries. If a text part follows a functionCall, the
+				// split creates an extra assistant turn between tool_use and tool_result,
+				// which Claude rejects with "tool_use ids were found without tool_result
+				// blocks immediately after".
 				if role == "model" {
 					partsResult := gjson.GetBytes(clientContentJSON, "parts")
 					if partsResult.IsArray() {
 						parts := partsResult.Array()
-						var thinkingParts []gjson.Result
-						var otherParts []gjson.Result
-						for _, part := range parts {
-							if part.Get("thought").Bool() {
-								thinkingParts = append(thinkingParts, part)
-							} else {
-								otherParts = append(otherParts, part)
-							}
-						}
-						if len(thinkingParts) > 0 {
-							firstPartIsThinking := parts[0].Get("thought").Bool()
-							if !firstPartIsThinking || len(thinkingParts) > 1 {
-								var newParts []interface{}
-								for _, p := range thinkingParts {
-									newParts = append(newParts, p.Value())
+						if len(parts) > 1 {
+							var thinkingParts []gjson.Result
+							var regularParts []gjson.Result
+							var functionCallParts []gjson.Result
+							for _, part := range parts {
+								if part.Get("thought").Bool() {
+									thinkingParts = append(thinkingParts, part)
+								} else if part.Get("functionCall").Exists() {
+									functionCallParts = append(functionCallParts, part)
+								} else {
+									regularParts = append(regularParts, part)
 								}
-								for _, p := range otherParts {
-									newParts = append(newParts, p.Value())
-								}
-								clientContentJSON, _ = sjson.SetBytes(clientContentJSON, "parts", newParts)
 							}
+							var newParts []interface{}
+							for _, p := range thinkingParts {
+								newParts = append(newParts, p.Value())
+							}
+							for _, p := range regularParts {
+								newParts = append(newParts, p.Value())
+							}
+							for _, p := range functionCallParts {
+								newParts = append(newParts, p.Value())
+							}
+							clientContentJSON, _ = sjson.SetBytes(clientContentJSON, "parts", newParts)
 						}
 					}
 				}

internal/translator/antigravity/claude/antigravity_claude_request_test.go

@@ -361,6 +361,167 @@ func TestConvertClaudeRequestToAntigravity_ReorderThinking(t *testing.T) {
 	}
 }
 
+func TestConvertClaudeRequestToAntigravity_ReorderTextAfterFunctionCall(t *testing.T) {
+	// Bug: text part after tool_use in an assistant message causes Antigravity
+	// to split at functionCall boundary, creating an extra assistant turn that
+	// breaks tool_use↔tool_result adjacency (upstream issue #989).
+	// Fix: reorder parts so functionCall comes last.
+	inputJSON := []byte(`{
+		"model": "claude-sonnet-4-5",
+		"messages": [
+			{
+				"role": "assistant",
+				"content": [
+					{"type": "text", "text": "Let me check..."},
+					{
+						"type": "tool_use",
+						"id": "call_abc",
+						"name": "Read",
+						"input": {"file": "test.go"}
+					},
+					{"type": "text", "text": "Reading the file now"}
+				]
+			},
+			{
+				"role": "user",
+				"content": [
+					{
+						"type": "tool_result",
+						"tool_use_id": "call_abc",
+						"content": "file content"
+					}
+				]
+			}
+		]
+	}`)
+
+	output := ConvertClaudeRequestToAntigravity("claude-sonnet-4-5", inputJSON, false)
+	outputStr := string(output)
+
+	parts := gjson.Get(outputStr, "request.contents.0.parts").Array()
+	if len(parts) != 3 {
+		t.Fatalf("Expected 3 parts, got %d", len(parts))
+	}
+
+	// Text parts should come before functionCall
+	if parts[0].Get("text").String() != "Let me check..." {
+		t.Errorf("Expected first text part first, got %s", parts[0].Raw)
+	}
+	if parts[1].Get("text").String() != "Reading the file now" {
+		t.Errorf("Expected second text part second, got %s", parts[1].Raw)
+	}
+	if !parts[2].Get("functionCall").Exists() {
+		t.Errorf("Expected functionCall last, got %s", parts[2].Raw)
+	}
+	if parts[2].Get("functionCall.name").String() != "Read" {
+		t.Errorf("Expected functionCall name 'Read', got '%s'", parts[2].Get("functionCall.name").String())
+	}
+}
+
+func TestConvertClaudeRequestToAntigravity_ReorderParallelFunctionCalls(t *testing.T) {
+	inputJSON := []byte(`{
+		"model": "claude-sonnet-4-5",
+		"messages": [
+			{
+				"role": "assistant",
+				"content": [
+					{"type": "text", "text": "Reading both files."},
+					{
+						"type": "tool_use",
+						"id": "call_1",
+						"name": "Read",
+						"input": {"file": "a.go"}
+					},
+					{"type": "text", "text": "And this one too."},
+					{
+						"type": "tool_use",
+						"id": "call_2",
+						"name": "Read",
+						"input": {"file": "b.go"}
+					}
+				]
+			}
+		]
+	}`)
+
+	output := ConvertClaudeRequestToAntigravity("claude-sonnet-4-5", inputJSON, false)
+	outputStr := string(output)
+
+	parts := gjson.Get(outputStr, "request.contents.0.parts").Array()
+	if len(parts) != 4 {
+		t.Fatalf("Expected 4 parts, got %d", len(parts))
+	}
+
+	if parts[0].Get("text").String() != "Reading both files." {
+		t.Errorf("Expected first text, got %s", parts[0].Raw)
+	}
+	if parts[1].Get("text").String() != "And this one too." {
+		t.Errorf("Expected second text, got %s", parts[1].Raw)
+	}
+	if parts[2].Get("functionCall.name").String() != "Read" || parts[2].Get("functionCall.id").String() != "call_1" {
+		t.Errorf("Expected fc1 third, got %s", parts[2].Raw)
+	}
+	if parts[3].Get("functionCall.name").String() != "Read" || parts[3].Get("functionCall.id").String() != "call_2" {
+		t.Errorf("Expected fc2 fourth, got %s", parts[3].Raw)
+	}
+}
+
+func TestConvertClaudeRequestToAntigravity_ReorderThinkingAndTextBeforeFunctionCall(t *testing.T) {
+	cache.ClearSignatureCache("")
+
+	validSignature := "abc123validSignature1234567890123456789012345678901234567890"
+	thinkingText := "Let me think about this..."
+
+	inputJSON := []byte(`{
+		"model": "claude-sonnet-4-5-thinking",
+		"messages": [
+			{
+				"role": "user",
+				"content": [{"type": "text", "text": "Hello"}]
+			},
+			{
+				"role": "assistant",
+				"content": [
+					{"type": "text", "text": "Before thinking"},
+					{"type": "thinking", "thinking": "` + thinkingText + `", "signature": "` + validSignature + `"},
+					{
+						"type": "tool_use",
+						"id": "call_xyz",
+						"name": "Bash",
+						"input": {"command": "ls"}
+					},
+					{"type": "text", "text": "After tool call"}
+				]
+			}
+		]
+	}`)
+
+	cache.CacheSignature("claude-sonnet-4-5-thinking", thinkingText, validSignature)
+
+	output := ConvertClaudeRequestToAntigravity("claude-sonnet-4-5-thinking", inputJSON, false)
+	outputStr := string(output)
+
+	// contents.1 = assistant message (contents.0 = user)
+	parts := gjson.Get(outputStr, "request.contents.1.parts").Array()
+	if len(parts) != 4 {
+		t.Fatalf("Expected 4 parts, got %d", len(parts))
+	}
+
+	// Order: thinking → text → text → functionCall
+	if !parts[0].Get("thought").Bool() {
+		t.Error("First part should be thinking")
+	}
+	if parts[1].Get("functionCall").Exists() || parts[1].Get("thought").Bool() {
+		t.Errorf("Second part should be text, got %s", parts[1].Raw)
+	}
+	if parts[2].Get("functionCall").Exists() || parts[2].Get("thought").Bool() {
+		t.Errorf("Third part should be text, got %s", parts[2].Raw)
+	}
+	if !parts[3].Get("functionCall").Exists() {
+		t.Errorf("Last part should be functionCall, got %s", parts[3].Raw)
+	}
+}
+
 func TestConvertClaudeRequestToAntigravity_ToolResult(t *testing.T) {
 	inputJSON := []byte(`{
 		"model": "claude-3-5-sonnet-20240620",

internal/translator/codex/openai/chat-completions/codex_openai_response.go

@@ -284,12 +284,12 @@ func ConvertCodexResponseToOpenAINonStream(_ context.Context, _ string, original
 	}
 
 	// Process the output array for content and function calls
+	var toolCalls [][]byte
 	outputResult := responseResult.Get("output")
 	if outputResult.IsArray() {
 		outputArray := outputResult.Array()
 		var contentText string
 		var reasoningText string
-		var toolCalls [][]byte
 
 		for _, outputItem := range outputArray {
 			outputType := outputItem.Get("type").String()
@@ -367,8 +367,12 @@ func ConvertCodexResponseToOpenAINonStream(_ context.Context, _ string, original
 	if statusResult := responseResult.Get("status"); statusResult.Exists() {
 		status := statusResult.String()
 		if status == "completed" {
-			template, _ = sjson.SetBytes(template, "choices.0.finish_reason", "stop")
-			template, _ = sjson.SetBytes(template, "choices.0.native_finish_reason", "stop")
+			finishReason := "stop"
+			if len(toolCalls) > 0 {
+				finishReason = "tool_calls"
+			}
+			template, _ = sjson.SetBytes(template, "choices.0.finish_reason", finishReason)
+			template, _ = sjson.SetBytes(template, "choices.0.native_finish_reason", finishReason)
 		}
 	}
 

internal/translator/gemini/claude/gemini_claude_request.go

@@ -6,7 +6,7 @@
 package claude
 
 import (
-	"bytes"
+	"fmt"
 	"strings"
 
 	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
@@ -31,8 +31,6 @@ const geminiClaudeThoughtSignature = "skip_thought_signature_validator"
 //   - []byte: The transformed request in Gemini CLI format.
 func ConvertClaudeRequestToGemini(modelName string, inputRawJSON []byte, _ bool) []byte {
 	rawJSON := inputRawJSON
-	rawJSON = bytes.Replace(rawJSON, []byte(`"url":{"type":"string","format":"uri",`), []byte(`"url":{"type":"string",`), -1)
-
 	// Build output Gemini CLI request JSON
 	out := []byte(`{"contents":[]}`)
 	out, _ = sjson.SetBytes(out, "model", modelName)
@@ -146,13 +144,37 @@ func ConvertClaudeRequestToGemini(modelName string, inputRawJSON []byte, _ bool)
 		})
 	}
 
+	// strip trailing model turn with unanswered function calls —
+	// Gemini returns empty responses when the last turn is a model
+	// functionCall with no corresponding user functionResponse.
+	contents := gjson.GetBytes(out, "contents")
+	if contents.Exists() && contents.IsArray() {
+		arr := contents.Array()
+		if len(arr) > 0 {
+			last := arr[len(arr)-1]
+			if last.Get("role").String() == "model" {
+				hasFC := false
+				last.Get("parts").ForEach(func(_, part gjson.Result) bool {
+					if part.Get("functionCall").Exists() {
+						hasFC = true
+						return false
+					}
+					return true
+				})
+				if hasFC {
+					out, _ = sjson.DeleteBytes(out, fmt.Sprintf("contents.%d", len(arr)-1))
+				}
+			}
+		}
+	}
+
 	// tools
 	if toolsResult := gjson.GetBytes(rawJSON, "tools"); toolsResult.IsArray() {
 		hasTools := false
 		toolsResult.ForEach(func(_, toolResult gjson.Result) bool {
 			inputSchemaResult := toolResult.Get("input_schema")
 			if inputSchemaResult.Exists() && inputSchemaResult.IsObject() {
-				inputSchema := inputSchemaResult.Raw
+				inputSchema := util.CleanJSONSchemaForGemini(inputSchemaResult.Raw)
 				tool := []byte(toolResult.Raw)
 				var err error
 				tool, err = sjson.DeleteBytes(tool, "input_schema")
@@ -168,6 +190,7 @@ func ConvertClaudeRequestToGemini(modelName string, inputRawJSON []byte, _ bool)
 				tool, _ = sjson.DeleteBytes(tool, "type")
 				tool, _ = sjson.DeleteBytes(tool, "cache_control")
 				tool, _ = sjson.DeleteBytes(tool, "defer_loading")
+				tool, _ = sjson.DeleteBytes(tool, "eager_input_streaming")
 				tool, _ = sjson.SetBytes(tool, "name", util.SanitizeFunctionName(gjson.GetBytes(tool, "name").String()))
 				if gjson.ValidBytes(tool) && gjson.ParseBytes(tool).IsObject() {
 					if !hasTools {

internal/translator/openai/openai/responses/openai_openai-responses_response.go

@@ -4,6 +4,7 @@ import (
 	"bytes"
 	"context"
 	"fmt"
+	"sort"
 	"strings"
 	"sync/atomic"
 	"time"
@@ -16,6 +17,7 @@ import (
 type oaiToResponsesStateReasoning struct {
 	ReasoningID   string
 	ReasoningData string
+	OutputIndex   int
 }
 type oaiToResponsesState struct {
 	Seq            int
@@ -29,16 +31,19 @@ type oaiToResponsesState struct {
 	MsgTextBuf   map[int]*strings.Builder
 	ReasoningBuf strings.Builder
 	Reasonings   []oaiToResponsesStateReasoning
-	FuncArgsBuf  map[int]*strings.Builder // index -> args
-	FuncNames    map[int]string           // index -> name
-	FuncCallIDs  map[int]string           // index -> call_id
+	FuncArgsBuf  map[string]*strings.Builder
+	FuncNames    map[string]string
+	FuncCallIDs  map[string]string
+	FuncOutputIx map[string]int
+	MsgOutputIx  map[int]int
+	NextOutputIx int
 	// message item state per output index
 	MsgItemAdded    map[int]bool // whether response.output_item.added emitted for message
 	MsgContentAdded map[int]bool // whether response.content_part.added emitted for message
 	MsgItemDone     map[int]bool // whether message done events were emitted
 	// function item done state
-	FuncArgsDone map[int]bool
-	FuncItemDone map[int]bool
+	FuncArgsDone map[string]bool
+	FuncItemDone map[string]bool
 	// usage aggregation
 	PromptTokens     int64
 	CachedTokens     int64
@@ -60,15 +65,17 @@ func emitRespEvent(event string, payload []byte) []byte {
 func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context, modelName string, originalRequestRawJSON, requestRawJSON, rawJSON []byte, param *any) [][]byte {
 	if *param == nil {
 		*param = &oaiToResponsesState{
-			FuncArgsBuf:     make(map[int]*strings.Builder),
-			FuncNames:       make(map[int]string),
-			FuncCallIDs:     make(map[int]string),
+			FuncArgsBuf:     make(map[string]*strings.Builder),
+			FuncNames:       make(map[string]string),
+			FuncCallIDs:     make(map[string]string),
+			FuncOutputIx:    make(map[string]int),
+			MsgOutputIx:     make(map[int]int),
 			MsgTextBuf:      make(map[int]*strings.Builder),
 			MsgItemAdded:    make(map[int]bool),
 			MsgContentAdded: make(map[int]bool),
 			MsgItemDone:     make(map[int]bool),
-			FuncArgsDone:    make(map[int]bool),
-			FuncItemDone:    make(map[int]bool),
+			FuncArgsDone:    make(map[string]bool),
+			FuncItemDone:    make(map[string]bool),
 			Reasonings:      make([]oaiToResponsesStateReasoning, 0),
 		}
 	}
@@ -125,6 +132,12 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 	}
 
 	nextSeq := func() int { st.Seq++; return st.Seq }
+	allocOutputIndex := func() int {
+		ix := st.NextOutputIx
+		st.NextOutputIx++
+		return ix
+	}
+	toolStateKey := func(outputIndex, toolIndex int) string { return fmt.Sprintf("%d:%d", outputIndex, toolIndex) }
 	var out [][]byte
 
 	if !st.Started {
@@ -135,14 +148,17 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 		st.ReasoningBuf.Reset()
 		st.ReasoningID = ""
 		st.ReasoningIndex = 0
-		st.FuncArgsBuf = make(map[int]*strings.Builder)
-		st.FuncNames = make(map[int]string)
-		st.FuncCallIDs = make(map[int]string)
+		st.FuncArgsBuf = make(map[string]*strings.Builder)
+		st.FuncNames = make(map[string]string)
+		st.FuncCallIDs = make(map[string]string)
+		st.FuncOutputIx = make(map[string]int)
+		st.MsgOutputIx = make(map[int]int)
+		st.NextOutputIx = 0
 		st.MsgItemAdded = make(map[int]bool)
 		st.MsgContentAdded = make(map[int]bool)
 		st.MsgItemDone = make(map[int]bool)
-		st.FuncArgsDone = make(map[int]bool)
-		st.FuncItemDone = make(map[int]bool)
+		st.FuncArgsDone = make(map[string]bool)
+		st.FuncItemDone = make(map[string]bool)
 		st.PromptTokens = 0
 		st.CachedTokens = 0
 		st.CompletionTokens = 0
@@ -185,7 +201,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 		outputItemDone, _ = sjson.SetBytes(outputItemDone, "item.summary.text", text)
 		out = append(out, emitRespEvent("response.output_item.done", outputItemDone))
 
-		st.Reasonings = append(st.Reasonings, oaiToResponsesStateReasoning{ReasoningID: st.ReasoningID, ReasoningData: text})
+		st.Reasonings = append(st.Reasonings, oaiToResponsesStateReasoning{ReasoningID: st.ReasoningID, ReasoningData: text, OutputIndex: st.ReasoningIndex})
 		st.ReasoningID = ""
 	}
 
@@ -201,10 +217,14 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 						stopReasoning(st.ReasoningBuf.String())
 						st.ReasoningBuf.Reset()
 					}
+					if _, exists := st.MsgOutputIx[idx]; !exists {
+						st.MsgOutputIx[idx] = allocOutputIndex()
+					}
+					msgOutputIndex := st.MsgOutputIx[idx]
 					if !st.MsgItemAdded[idx] {
 						item := []byte(`{"type":"response.output_item.added","sequence_number":0,"output_index":0,"item":{"id":"","type":"message","status":"in_progress","content":[],"role":"assistant"}}`)
 						item, _ = sjson.SetBytes(item, "sequence_number", nextSeq())
-						item, _ = sjson.SetBytes(item, "output_index", idx)
+						item, _ = sjson.SetBytes(item, "output_index", msgOutputIndex)
 						item, _ = sjson.SetBytes(item, "item.id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
 						out = append(out, emitRespEvent("response.output_item.added", item))
 						st.MsgItemAdded[idx] = true
@@ -213,7 +233,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 						part := []byte(`{"type":"response.content_part.added","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"part":{"type":"output_text","annotations":[],"logprobs":[],"text":""}}`)
 						part, _ = sjson.SetBytes(part, "sequence_number", nextSeq())
 						part, _ = sjson.SetBytes(part, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
-						part, _ = sjson.SetBytes(part, "output_index", idx)
+						part, _ = sjson.SetBytes(part, "output_index", msgOutputIndex)
 						part, _ = sjson.SetBytes(part, "content_index", 0)
 						out = append(out, emitRespEvent("response.content_part.added", part))
 						st.MsgContentAdded[idx] = true
@@ -222,7 +242,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 					msg := []byte(`{"type":"response.output_text.delta","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"delta":"","logprobs":[]}`)
 					msg, _ = sjson.SetBytes(msg, "sequence_number", nextSeq())
 					msg, _ = sjson.SetBytes(msg, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
-					msg, _ = sjson.SetBytes(msg, "output_index", idx)
+					msg, _ = sjson.SetBytes(msg, "output_index", msgOutputIndex)
 					msg, _ = sjson.SetBytes(msg, "content_index", 0)
 					msg, _ = sjson.SetBytes(msg, "delta", c.String())
 					out = append(out, emitRespEvent("response.output_text.delta", msg))
@@ -238,10 +258,10 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 					// On first appearance, add reasoning item and part
 					if st.ReasoningID == "" {
 						st.ReasoningID = fmt.Sprintf("rs_%s_%d", st.ResponseID, idx)
-						st.ReasoningIndex = idx
+						st.ReasoningIndex = allocOutputIndex()
 						item := []byte(`{"type":"response.output_item.added","sequence_number":0,"output_index":0,"item":{"id":"","type":"reasoning","status":"in_progress","summary":[]}}`)
 						item, _ = sjson.SetBytes(item, "sequence_number", nextSeq())
-						item, _ = sjson.SetBytes(item, "output_index", idx)
+						item, _ = sjson.SetBytes(item, "output_index", st.ReasoningIndex)
 						item, _ = sjson.SetBytes(item, "item.id", st.ReasoningID)
 						out = append(out, emitRespEvent("response.output_item.added", item))
 						part := []byte(`{"type":"response.reasoning_summary_part.added","sequence_number":0,"item_id":"","output_index":0,"summary_index":0,"part":{"type":"summary_text","text":""}}`)
@@ -269,6 +289,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 					// Before emitting any function events, if a message is open for this index,
 					// close its text/content to match Codex expected ordering.
 					if st.MsgItemAdded[idx] && !st.MsgItemDone[idx] {
+						msgOutputIndex := st.MsgOutputIx[idx]
 						fullText := ""
 						if b := st.MsgTextBuf[idx]; b != nil {
 							fullText = b.String()
@@ -276,7 +297,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 						done := []byte(`{"type":"response.output_text.done","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"text":"","logprobs":[]}`)
 						done, _ = sjson.SetBytes(done, "sequence_number", nextSeq())
 						done, _ = sjson.SetBytes(done, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
-						done, _ = sjson.SetBytes(done, "output_index", idx)
+						done, _ = sjson.SetBytes(done, "output_index", msgOutputIndex)
 						done, _ = sjson.SetBytes(done, "content_index", 0)
 						done, _ = sjson.SetBytes(done, "text", fullText)
 						out = append(out, emitRespEvent("response.output_text.done", done))
@@ -284,69 +305,72 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 						partDone := []byte(`{"type":"response.content_part.done","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"part":{"type":"output_text","annotations":[],"logprobs":[],"text":""}}`)
 						partDone, _ = sjson.SetBytes(partDone, "sequence_number", nextSeq())
 						partDone, _ = sjson.SetBytes(partDone, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
-						partDone, _ = sjson.SetBytes(partDone, "output_index", idx)
+						partDone, _ = sjson.SetBytes(partDone, "output_index", msgOutputIndex)
 						partDone, _ = sjson.SetBytes(partDone, "content_index", 0)
 						partDone, _ = sjson.SetBytes(partDone, "part.text", fullText)
 						out = append(out, emitRespEvent("response.content_part.done", partDone))
 
 						itemDone := []byte(`{"type":"response.output_item.done","sequence_number":0,"output_index":0,"item":{"id":"","type":"message","status":"completed","content":[{"type":"output_text","annotations":[],"logprobs":[],"text":""}],"role":"assistant"}}`)
 						itemDone, _ = sjson.SetBytes(itemDone, "sequence_number", nextSeq())
-						itemDone, _ = sjson.SetBytes(itemDone, "output_index", idx)
+						itemDone, _ = sjson.SetBytes(itemDone, "output_index", msgOutputIndex)
 						itemDone, _ = sjson.SetBytes(itemDone, "item.id", fmt.Sprintf("msg_%s_%d", st.ResponseID, idx))
 						itemDone, _ = sjson.SetBytes(itemDone, "item.content.0.text", fullText)
 						out = append(out, emitRespEvent("response.output_item.done", itemDone))
 						st.MsgItemDone[idx] = true
 					}
 
-					// Only emit item.added once per tool call and preserve call_id across chunks.
-					newCallID := tcs.Get("0.id").String()
-					nameChunk := tcs.Get("0.function.name").String()
-					if nameChunk != "" {
-						st.FuncNames[idx] = nameChunk
-					}
-					existingCallID := st.FuncCallIDs[idx]
-					effectiveCallID := existingCallID
-					shouldEmitItem := false
-					if existingCallID == "" && newCallID != "" {
-						// First time seeing a valid call_id for this index
-						effectiveCallID = newCallID
-						st.FuncCallIDs[idx] = newCallID
-						shouldEmitItem = true
-					}
-
-					if shouldEmitItem && effectiveCallID != "" {
-						o := []byte(`{"type":"response.output_item.added","sequence_number":0,"output_index":0,"item":{"id":"","type":"function_call","status":"in_progress","arguments":"","call_id":"","name":""}}`)
-						o, _ = sjson.SetBytes(o, "sequence_number", nextSeq())
-						o, _ = sjson.SetBytes(o, "output_index", idx)
-						o, _ = sjson.SetBytes(o, "item.id", fmt.Sprintf("fc_%s", effectiveCallID))
-						o, _ = sjson.SetBytes(o, "item.call_id", effectiveCallID)
-						name := st.FuncNames[idx]
-						o, _ = sjson.SetBytes(o, "item.name", name)
-						out = append(out, emitRespEvent("response.output_item.added", o))
-					}
-
-					// Ensure args buffer exists for this index
-					if st.FuncArgsBuf[idx] == nil {
-						st.FuncArgsBuf[idx] = &strings.Builder{}
-					}
-
-					// Append arguments delta if available and we have a valid call_id to reference
-					if args := tcs.Get("0.function.arguments"); args.Exists() && args.String() != "" {
-						// Prefer an already known call_id; fall back to newCallID if first time
-						refCallID := st.FuncCallIDs[idx]
-						if refCallID == "" {
-							refCallID = newCallID
+					tcs.ForEach(func(_, tc gjson.Result) bool {
+						toolIndex := int(tc.Get("index").Int())
+						key := toolStateKey(idx, toolIndex)
+						newCallID := tc.Get("id").String()
+						nameChunk := tc.Get("function.name").String()
+						if nameChunk != "" {
+							st.FuncNames[key] = nameChunk
 						}
-						if refCallID != "" {
-							ad := []byte(`{"type":"response.function_call_arguments.delta","sequence_number":0,"item_id":"","output_index":0,"delta":""}`)
-							ad, _ = sjson.SetBytes(ad, "sequence_number", nextSeq())
-							ad, _ = sjson.SetBytes(ad, "item_id", fmt.Sprintf("fc_%s", refCallID))
-							ad, _ = sjson.SetBytes(ad, "output_index", idx)
-							ad, _ = sjson.SetBytes(ad, "delta", args.String())
-							out = append(out, emitRespEvent("response.function_call_arguments.delta", ad))
+
+						existingCallID := st.FuncCallIDs[key]
+						effectiveCallID := existingCallID
+						shouldEmitItem := false
+						if existingCallID == "" && newCallID != "" {
+							effectiveCallID = newCallID
+							st.FuncCallIDs[key] = newCallID
+							st.FuncOutputIx[key] = allocOutputIndex()
+							shouldEmitItem = true
 						}
-						st.FuncArgsBuf[idx].WriteString(args.String())
-					}
+
+						if shouldEmitItem && effectiveCallID != "" {
+							outputIndex := st.FuncOutputIx[key]
+							o := []byte(`{"type":"response.output_item.added","sequence_number":0,"output_index":0,"item":{"id":"","type":"function_call","status":"in_progress","arguments":"","call_id":"","name":""}}`)
+							o, _ = sjson.SetBytes(o, "sequence_number", nextSeq())
+							o, _ = sjson.SetBytes(o, "output_index", outputIndex)
+							o, _ = sjson.SetBytes(o, "item.id", fmt.Sprintf("fc_%s", effectiveCallID))
+							o, _ = sjson.SetBytes(o, "item.call_id", effectiveCallID)
+							o, _ = sjson.SetBytes(o, "item.name", st.FuncNames[key])
+							out = append(out, emitRespEvent("response.output_item.added", o))
+						}
+
+						if st.FuncArgsBuf[key] == nil {
+							st.FuncArgsBuf[key] = &strings.Builder{}
+						}
+
+						if args := tc.Get("function.arguments"); args.Exists() && args.String() != "" {
+							refCallID := st.FuncCallIDs[key]
+							if refCallID == "" {
+								refCallID = newCallID
+							}
+							if refCallID != "" {
+								outputIndex := st.FuncOutputIx[key]
+								ad := []byte(`{"type":"response.function_call_arguments.delta","sequence_number":0,"item_id":"","output_index":0,"delta":""}`)
+								ad, _ = sjson.SetBytes(ad, "sequence_number", nextSeq())
+								ad, _ = sjson.SetBytes(ad, "item_id", fmt.Sprintf("fc_%s", refCallID))
+								ad, _ = sjson.SetBytes(ad, "output_index", outputIndex)
+								ad, _ = sjson.SetBytes(ad, "delta", args.String())
+								out = append(out, emitRespEvent("response.function_call_arguments.delta", ad))
+							}
+							st.FuncArgsBuf[key].WriteString(args.String())
+						}
+						return true
+					})
 				}
 			}
 
@@ -360,15 +384,10 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 					for i := range st.MsgItemAdded {
 						idxs = append(idxs, i)
 					}
-					for i := 0; i < len(idxs); i++ {
-						for j := i + 1; j < len(idxs); j++ {
-							if idxs[j] < idxs[i] {
-								idxs[i], idxs[j] = idxs[j], idxs[i]
-							}
-						}
-					}
+					sort.Slice(idxs, func(i, j int) bool { return st.MsgOutputIx[idxs[i]] < st.MsgOutputIx[idxs[j]] })
 					for _, i := range idxs {
 						if st.MsgItemAdded[i] && !st.MsgItemDone[i] {
+							msgOutputIndex := st.MsgOutputIx[i]
 							fullText := ""
 							if b := st.MsgTextBuf[i]; b != nil {
 								fullText = b.String()
@@ -376,7 +395,7 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 							done := []byte(`{"type":"response.output_text.done","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"text":"","logprobs":[]}`)
 							done, _ = sjson.SetBytes(done, "sequence_number", nextSeq())
 							done, _ = sjson.SetBytes(done, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, i))
-							done, _ = sjson.SetBytes(done, "output_index", i)
+							done, _ = sjson.SetBytes(done, "output_index", msgOutputIndex)
 							done, _ = sjson.SetBytes(done, "content_index", 0)
 							done, _ = sjson.SetBytes(done, "text", fullText)
 							out = append(out, emitRespEvent("response.output_text.done", done))
@@ -384,14 +403,14 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 							partDone := []byte(`{"type":"response.content_part.done","sequence_number":0,"item_id":"","output_index":0,"content_index":0,"part":{"type":"output_text","annotations":[],"logprobs":[],"text":""}}`)
 							partDone, _ = sjson.SetBytes(partDone, "sequence_number", nextSeq())
 							partDone, _ = sjson.SetBytes(partDone, "item_id", fmt.Sprintf("msg_%s_%d", st.ResponseID, i))
-							partDone, _ = sjson.SetBytes(partDone, "output_index", i)
+							partDone, _ = sjson.SetBytes(partDone, "output_index", msgOutputIndex)
 							partDone, _ = sjson.SetBytes(partDone, "content_index", 0)
 							partDone, _ = sjson.SetBytes(partDone, "part.text", fullText)
 							out = append(out, emitRespEvent("response.content_part.done", partDone))
 
 							itemDone := []byte(`{"type":"response.output_item.done","sequence_number":0,"output_index":0,"item":{"id":"","type":"message","status":"completed","content":[{"type":"output_text","annotations":[],"logprobs":[],"text":""}],"role":"assistant"}}`)
 							itemDone, _ = sjson.SetBytes(itemDone, "sequence_number", nextSeq())
-							itemDone, _ = sjson.SetBytes(itemDone, "output_index", i)
+							itemDone, _ = sjson.SetBytes(itemDone, "output_index", msgOutputIndex)
 							itemDone, _ = sjson.SetBytes(itemDone, "item.id", fmt.Sprintf("msg_%s_%d", st.ResponseID, i))
 							itemDone, _ = sjson.SetBytes(itemDone, "item.content.0.text", fullText)
 							out = append(out, emitRespEvent("response.output_item.done", itemDone))
@@ -407,43 +426,42 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 
 				// Emit function call done events for any active function calls
 				if len(st.FuncCallIDs) > 0 {
-					idxs := make([]int, 0, len(st.FuncCallIDs))
-					for i := range st.FuncCallIDs {
-						idxs = append(idxs, i)
+					keys := make([]string, 0, len(st.FuncCallIDs))
+					for key := range st.FuncCallIDs {
+						keys = append(keys, key)
 					}
-					for i := 0; i < len(idxs); i++ {
-						for j := i + 1; j < len(idxs); j++ {
-							if idxs[j] < idxs[i] {
-								idxs[i], idxs[j] = idxs[j], idxs[i]
-							}
-						}
-					}
-					for _, i := range idxs {
-						callID := st.FuncCallIDs[i]
-						if callID == "" || st.FuncItemDone[i] {
+					sort.Slice(keys, func(i, j int) bool {
+						left := st.FuncOutputIx[keys[i]]
+						right := st.FuncOutputIx[keys[j]]
+						return left < right || (left == right && keys[i] < keys[j])
+					})
+					for _, key := range keys {
+						callID := st.FuncCallIDs[key]
+						if callID == "" || st.FuncItemDone[key] {
 							continue
 						}
+						outputIndex := st.FuncOutputIx[key]
 						args := "{}"
-						if b := st.FuncArgsBuf[i]; b != nil && b.Len() > 0 {
+						if b := st.FuncArgsBuf[key]; b != nil && b.Len() > 0 {
 							args = b.String()
 						}
 						fcDone := []byte(`{"type":"response.function_call_arguments.done","sequence_number":0,"item_id":"","output_index":0,"arguments":""}`)
 						fcDone, _ = sjson.SetBytes(fcDone, "sequence_number", nextSeq())
 						fcDone, _ = sjson.SetBytes(fcDone, "item_id", fmt.Sprintf("fc_%s", callID))
-						fcDone, _ = sjson.SetBytes(fcDone, "output_index", i)
+						fcDone, _ = sjson.SetBytes(fcDone, "output_index", outputIndex)
 						fcDone, _ = sjson.SetBytes(fcDone, "arguments", args)
 						out = append(out, emitRespEvent("response.function_call_arguments.done", fcDone))
 
 						itemDone := []byte(`{"type":"response.output_item.done","sequence_number":0,"output_index":0,"item":{"id":"","type":"function_call","status":"completed","arguments":"","call_id":"","name":""}}`)
 						itemDone, _ = sjson.SetBytes(itemDone, "sequence_number", nextSeq())
-						itemDone, _ = sjson.SetBytes(itemDone, "output_index", i)
+						itemDone, _ = sjson.SetBytes(itemDone, "output_index", outputIndex)
 						itemDone, _ = sjson.SetBytes(itemDone, "item.id", fmt.Sprintf("fc_%s", callID))
 						itemDone, _ = sjson.SetBytes(itemDone, "item.arguments", args)
 						itemDone, _ = sjson.SetBytes(itemDone, "item.call_id", callID)
-						itemDone, _ = sjson.SetBytes(itemDone, "item.name", st.FuncNames[i])
+						itemDone, _ = sjson.SetBytes(itemDone, "item.name", st.FuncNames[key])
 						out = append(out, emitRespEvent("response.output_item.done", itemDone))
-						st.FuncItemDone[i] = true
-						st.FuncArgsDone[i] = true
+						st.FuncItemDone[key] = true
+						st.FuncArgsDone[key] = true
 					}
 				}
 				completed := []byte(`{"type":"response.completed","sequence_number":0,"response":{"id":"","object":"response","created_at":0,"status":"completed","background":false,"error":null}}`)
@@ -516,28 +534,21 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 				}
 				// Build response.output using aggregated buffers
 				outputsWrapper := []byte(`{"arr":[]}`)
+				type completedOutputItem struct {
+					index int
+					raw   []byte
+				}
+				outputItems := make([]completedOutputItem, 0, len(st.Reasonings)+len(st.MsgItemAdded)+len(st.FuncArgsBuf))
 				if len(st.Reasonings) > 0 {
 					for _, r := range st.Reasonings {
 						item := []byte(`{"id":"","type":"reasoning","summary":[{"type":"summary_text","text":""}]}`)
 						item, _ = sjson.SetBytes(item, "id", r.ReasoningID)
 						item, _ = sjson.SetBytes(item, "summary.0.text", r.ReasoningData)
-						outputsWrapper, _ = sjson.SetRawBytes(outputsWrapper, "arr.-1", item)
+						outputItems = append(outputItems, completedOutputItem{index: r.OutputIndex, raw: item})
 					}
 				}
-				// Append message items in ascending index order
 				if len(st.MsgItemAdded) > 0 {
-					midxs := make([]int, 0, len(st.MsgItemAdded))
 					for i := range st.MsgItemAdded {
-						midxs = append(midxs, i)
-					}
-					for i := 0; i < len(midxs); i++ {
-						for j := i + 1; j < len(midxs); j++ {
-							if midxs[j] < midxs[i] {
-								midxs[i], midxs[j] = midxs[j], midxs[i]
-							}
-						}
-					}
-					for _, i := range midxs {
 						txt := ""
 						if b := st.MsgTextBuf[i]; b != nil {
 							txt = b.String()
@@ -545,37 +556,29 @@ func ConvertOpenAIChatCompletionsResponseToOpenAIResponses(ctx context.Context,
 						item := []byte(`{"id":"","type":"message","status":"completed","content":[{"type":"output_text","annotations":[],"logprobs":[],"text":""}],"role":"assistant"}`)
 						item, _ = sjson.SetBytes(item, "id", fmt.Sprintf("msg_%s_%d", st.ResponseID, i))
 						item, _ = sjson.SetBytes(item, "content.0.text", txt)
-						outputsWrapper, _ = sjson.SetRawBytes(outputsWrapper, "arr.-1", item)
+						outputItems = append(outputItems, completedOutputItem{index: st.MsgOutputIx[i], raw: item})
 					}
 				}
 				if len(st.FuncArgsBuf) > 0 {
-					idxs := make([]int, 0, len(st.FuncArgsBuf))
-					for i := range st.FuncArgsBuf {
-						idxs = append(idxs, i)
-					}
-					// small-N sort without extra imports
-					for i := 0; i < len(idxs); i++ {
-						for j := i + 1; j < len(idxs); j++ {
-							if idxs[j] < idxs[i] {
-								idxs[i], idxs[j] = idxs[j], idxs[i]
-							}
-						}
-					}
-					for _, i := range idxs {
+					for key := range st.FuncArgsBuf {
 						args := ""
-						if b := st.FuncArgsBuf[i]; b != nil {
+						if b := st.FuncArgsBuf[key]; b != nil {
 							args = b.String()
 						}
-						callID := st.FuncCallIDs[i]
-						name := st.FuncNames[i]
+						callID := st.FuncCallIDs[key]
+						name := st.FuncNames[key]
 						item := []byte(`{"id":"","type":"function_call","status":"completed","arguments":"","call_id":"","name":""}`)
 						item, _ = sjson.SetBytes(item, "id", fmt.Sprintf("fc_%s", callID))
 						item, _ = sjson.SetBytes(item, "arguments", args)
 						item, _ = sjson.SetBytes(item, "call_id", callID)
 						item, _ = sjson.SetBytes(item, "name", name)
-						outputsWrapper, _ = sjson.SetRawBytes(outputsWrapper, "arr.-1", item)
+						outputItems = append(outputItems, completedOutputItem{index: st.FuncOutputIx[key], raw: item})
 					}
 				}
+				sort.Slice(outputItems, func(i, j int) bool { return outputItems[i].index < outputItems[j].index })
+				for _, item := range outputItems {
+					outputsWrapper, _ = sjson.SetRawBytes(outputsWrapper, "arr.-1", item.raw)
+				}
 				if gjson.GetBytes(outputsWrapper, "arr.#").Int() > 0 {
 					completed, _ = sjson.SetRawBytes(completed, "response.output", []byte(gjson.GetBytes(outputsWrapper, "arr").Raw))
 				}

internal/translator/openai/openai/responses/openai_openai-responses_response_test.go

@@ -0,0 +1,305 @@
+package responses
+
+import (
+	"context"
+	"strings"
+	"testing"
+
+	"github.com/tidwall/gjson"
+)
+
+func parseOpenAIResponsesSSEEvent(t *testing.T, chunk []byte) (string, gjson.Result) {
+	t.Helper()
+
+	lines := strings.Split(string(chunk), "\n")
+	if len(lines) < 2 {
+		t.Fatalf("unexpected SSE chunk: %q", chunk)
+	}
+
+	event := strings.TrimSpace(strings.TrimPrefix(lines[0], "event:"))
+	dataLine := strings.TrimSpace(strings.TrimPrefix(lines[1], "data:"))
+	if !gjson.Valid(dataLine) {
+		t.Fatalf("invalid SSE data JSON: %q", dataLine)
+	}
+	return event, gjson.Parse(dataLine)
+}
+
+func TestConvertOpenAIChatCompletionsResponseToOpenAIResponses_MultipleToolCallsRemainSeparate(t *testing.T) {
+	in := []string{
+		`data: {"id":"resp_test","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":0,"id":"call_read","type":"function","function":{"name":"read","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_test","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":0,"function":{"arguments":"{\"filePath\":\"C:\\\\repo\",\"limit\":400,\"offset\":1}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_test","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":1,"id":"call_glob","type":"function","function":{"name":"glob","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_test","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":1,"function":{"arguments":"{\"path\":\"C:\\\\repo\",\"pattern\":\"*.{yml,yaml}\"}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_test","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":null},"finish_reason":"tool_calls"}],"usage":{"completion_tokens":10,"total_tokens":20,"prompt_tokens":10}}`,
+	}
+
+	request := []byte(`{"model":"gpt-5.4","tool_choice":"auto","parallel_tool_calls":true}`)
+
+	var param any
+	var out [][]byte
+	for _, line := range in {
+		out = append(out, ConvertOpenAIChatCompletionsResponseToOpenAIResponses(context.Background(), "model", request, request, []byte(line), &param)...)
+	}
+
+	addedNames := map[string]string{}
+	doneArgs := map[string]string{}
+	doneNames := map[string]string{}
+	outputItems := map[string]gjson.Result{}
+
+	for _, chunk := range out {
+		ev, data := parseOpenAIResponsesSSEEvent(t, chunk)
+		switch ev {
+		case "response.output_item.added":
+			if data.Get("item.type").String() != "function_call" {
+				continue
+			}
+			addedNames[data.Get("item.call_id").String()] = data.Get("item.name").String()
+		case "response.output_item.done":
+			if data.Get("item.type").String() != "function_call" {
+				continue
+			}
+			callID := data.Get("item.call_id").String()
+			doneArgs[callID] = data.Get("item.arguments").String()
+			doneNames[callID] = data.Get("item.name").String()
+		case "response.completed":
+			output := data.Get("response.output")
+			for _, item := range output.Array() {
+				if item.Get("type").String() == "function_call" {
+					outputItems[item.Get("call_id").String()] = item
+				}
+			}
+		}
+	}
+
+	if len(addedNames) != 2 {
+		t.Fatalf("expected 2 function_call added events, got %d", len(addedNames))
+	}
+	if len(doneArgs) != 2 {
+		t.Fatalf("expected 2 function_call done events, got %d", len(doneArgs))
+	}
+
+	if addedNames["call_read"] != "read" {
+		t.Fatalf("unexpected added name for call_read: %q", addedNames["call_read"])
+	}
+	if addedNames["call_glob"] != "glob" {
+		t.Fatalf("unexpected added name for call_glob: %q", addedNames["call_glob"])
+	}
+
+	if !gjson.Valid(doneArgs["call_read"]) {
+		t.Fatalf("invalid JSON args for call_read: %q", doneArgs["call_read"])
+	}
+	if !gjson.Valid(doneArgs["call_glob"]) {
+		t.Fatalf("invalid JSON args for call_glob: %q", doneArgs["call_glob"])
+	}
+	if strings.Contains(doneArgs["call_read"], "}{") {
+		t.Fatalf("call_read args were concatenated: %q", doneArgs["call_read"])
+	}
+	if strings.Contains(doneArgs["call_glob"], "}{") {
+		t.Fatalf("call_glob args were concatenated: %q", doneArgs["call_glob"])
+	}
+
+	if doneNames["call_read"] != "read" {
+		t.Fatalf("unexpected done name for call_read: %q", doneNames["call_read"])
+	}
+	if doneNames["call_glob"] != "glob" {
+		t.Fatalf("unexpected done name for call_glob: %q", doneNames["call_glob"])
+	}
+
+	if got := gjson.Get(doneArgs["call_read"], "filePath").String(); got != `C:\repo` {
+		t.Fatalf("unexpected filePath for call_read: %q", got)
+	}
+	if got := gjson.Get(doneArgs["call_glob"], "path").String(); got != `C:\repo` {
+		t.Fatalf("unexpected path for call_glob: %q", got)
+	}
+	if got := gjson.Get(doneArgs["call_glob"], "pattern").String(); got != "*.{yml,yaml}" {
+		t.Fatalf("unexpected pattern for call_glob: %q", got)
+	}
+
+	if len(outputItems) != 2 {
+		t.Fatalf("expected 2 function_call items in response.output, got %d", len(outputItems))
+	}
+	if outputItems["call_read"].Get("name").String() != "read" {
+		t.Fatalf("unexpected response.output name for call_read: %q", outputItems["call_read"].Get("name").String())
+	}
+	if outputItems["call_glob"].Get("name").String() != "glob" {
+		t.Fatalf("unexpected response.output name for call_glob: %q", outputItems["call_glob"].Get("name").String())
+	}
+}
+
+func TestConvertOpenAIChatCompletionsResponseToOpenAIResponses_MultiChoiceToolCallsUseDistinctOutputIndexes(t *testing.T) {
+	in := []string{
+		`data: {"id":"resp_multi_choice","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":0,"id":"call_choice0","type":"function","function":{"name":"glob","arguments":""}}]},"finish_reason":null},{"index":1,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":0,"id":"call_choice1","type":"function","function":{"name":"read","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_multi_choice","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":0,"function":{"arguments":"{\"path\":\"C:\\\\repo\",\"pattern\":\"*.go\"}"}}]},"finish_reason":null},{"index":1,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":0,"function":{"arguments":"{\"filePath\":\"C:\\\\repo\\\\README.md\",\"limit\":20,\"offset\":1}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_multi_choice","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":null},"finish_reason":"tool_calls"},{"index":1,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":null},"finish_reason":"tool_calls"}],"usage":{"completion_tokens":10,"total_tokens":20,"prompt_tokens":10}}`,
+	}
+
+	request := []byte(`{"model":"gpt-5.4","tool_choice":"auto","parallel_tool_calls":true}`)
+
+	var param any
+	var out [][]byte
+	for _, line := range in {
+		out = append(out, ConvertOpenAIChatCompletionsResponseToOpenAIResponses(context.Background(), "model", request, request, []byte(line), &param)...)
+	}
+
+	type fcEvent struct {
+		outputIndex int64
+		name        string
+		arguments   string
+	}
+
+	added := map[string]fcEvent{}
+	done := map[string]fcEvent{}
+
+	for _, chunk := range out {
+		ev, data := parseOpenAIResponsesSSEEvent(t, chunk)
+		switch ev {
+		case "response.output_item.added":
+			if data.Get("item.type").String() != "function_call" {
+				continue
+			}
+			callID := data.Get("item.call_id").String()
+			added[callID] = fcEvent{
+				outputIndex: data.Get("output_index").Int(),
+				name:        data.Get("item.name").String(),
+			}
+		case "response.output_item.done":
+			if data.Get("item.type").String() != "function_call" {
+				continue
+			}
+			callID := data.Get("item.call_id").String()
+			done[callID] = fcEvent{
+				outputIndex: data.Get("output_index").Int(),
+				name:        data.Get("item.name").String(),
+				arguments:   data.Get("item.arguments").String(),
+			}
+		}
+	}
+
+	if len(added) != 2 {
+		t.Fatalf("expected 2 function_call added events, got %d", len(added))
+	}
+	if len(done) != 2 {
+		t.Fatalf("expected 2 function_call done events, got %d", len(done))
+	}
+
+	if added["call_choice0"].name != "glob" {
+		t.Fatalf("unexpected added name for call_choice0: %q", added["call_choice0"].name)
+	}
+	if added["call_choice1"].name != "read" {
+		t.Fatalf("unexpected added name for call_choice1: %q", added["call_choice1"].name)
+	}
+	if added["call_choice0"].outputIndex == added["call_choice1"].outputIndex {
+		t.Fatalf("expected distinct output indexes for different choices, both got %d", added["call_choice0"].outputIndex)
+	}
+
+	if !gjson.Valid(done["call_choice0"].arguments) {
+		t.Fatalf("invalid JSON args for call_choice0: %q", done["call_choice0"].arguments)
+	}
+	if !gjson.Valid(done["call_choice1"].arguments) {
+		t.Fatalf("invalid JSON args for call_choice1: %q", done["call_choice1"].arguments)
+	}
+	if done["call_choice0"].outputIndex == done["call_choice1"].outputIndex {
+		t.Fatalf("expected distinct done output indexes for different choices, both got %d", done["call_choice0"].outputIndex)
+	}
+	if done["call_choice0"].name != "glob" {
+		t.Fatalf("unexpected done name for call_choice0: %q", done["call_choice0"].name)
+	}
+	if done["call_choice1"].name != "read" {
+		t.Fatalf("unexpected done name for call_choice1: %q", done["call_choice1"].name)
+	}
+}
+
+func TestConvertOpenAIChatCompletionsResponseToOpenAIResponses_MixedMessageAndToolUseDistinctOutputIndexes(t *testing.T) {
+	in := []string{
+		`data: {"id":"resp_mixed","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":"hello","reasoning_content":null,"tool_calls":null},"finish_reason":null},{"index":1,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":0,"id":"call_choice1","type":"function","function":{"name":"read","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_mixed","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":null},"finish_reason":"stop"},{"index":1,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":0,"function":{"arguments":"{\"filePath\":\"C:\\\\repo\\\\README.md\",\"limit\":20,\"offset\":1}"}}]},"finish_reason":"tool_calls"}],"usage":{"completion_tokens":10,"total_tokens":20,"prompt_tokens":10}}`,
+	}
+
+	request := []byte(`{"model":"gpt-5.4","tool_choice":"auto","parallel_tool_calls":true}`)
+
+	var param any
+	var out [][]byte
+	for _, line := range in {
+		out = append(out, ConvertOpenAIChatCompletionsResponseToOpenAIResponses(context.Background(), "model", request, request, []byte(line), &param)...)
+	}
+
+	var messageOutputIndex int64 = -1
+	var toolOutputIndex int64 = -1
+
+	for _, chunk := range out {
+		ev, data := parseOpenAIResponsesSSEEvent(t, chunk)
+		if ev != "response.output_item.added" {
+			continue
+		}
+		switch data.Get("item.type").String() {
+		case "message":
+			if data.Get("item.id").String() == "msg_resp_mixed_0" {
+				messageOutputIndex = data.Get("output_index").Int()
+			}
+		case "function_call":
+			if data.Get("item.call_id").String() == "call_choice1" {
+				toolOutputIndex = data.Get("output_index").Int()
+			}
+		}
+	}
+
+	if messageOutputIndex < 0 {
+		t.Fatal("did not find message output index")
+	}
+	if toolOutputIndex < 0 {
+		t.Fatal("did not find tool output index")
+	}
+	if messageOutputIndex == toolOutputIndex {
+		t.Fatalf("expected distinct output indexes for message and tool call, both got %d", messageOutputIndex)
+	}
+}
+
+func TestConvertOpenAIChatCompletionsResponseToOpenAIResponses_FunctionCallDoneAndCompletedOutputStayAscending(t *testing.T) {
+	in := []string{
+		`data: {"id":"resp_order","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":0,"id":"call_glob","type":"function","function":{"name":"glob","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_order","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":0,"function":{"arguments":"{\"path\":\"C:\\\\repo\",\"pattern\":\"*.go\"}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_order","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":"assistant","content":null,"reasoning_content":null,"tool_calls":[{"index":1,"id":"call_read","type":"function","function":{"name":"read","arguments":""}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_order","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":[{"index":1,"function":{"arguments":"{\"filePath\":\"C:\\\\repo\\\\README.md\",\"limit\":20,\"offset\":1}"}}]},"finish_reason":null}]}`,
+		`data: {"id":"resp_order","object":"chat.completion.chunk","created":1773896263,"model":"model","choices":[{"index":0,"delta":{"role":null,"content":null,"reasoning_content":null,"tool_calls":null},"finish_reason":"tool_calls"}],"usage":{"completion_tokens":10,"total_tokens":20,"prompt_tokens":10}}`,
+	}
+
+	request := []byte(`{"model":"gpt-5.4","tool_choice":"auto","parallel_tool_calls":true}`)
+
+	var param any
+	var out [][]byte
+	for _, line := range in {
+		out = append(out, ConvertOpenAIChatCompletionsResponseToOpenAIResponses(context.Background(), "model", request, request, []byte(line), &param)...)
+	}
+
+	var doneIndexes []int64
+	var completedOrder []string
+
+	for _, chunk := range out {
+		ev, data := parseOpenAIResponsesSSEEvent(t, chunk)
+		switch ev {
+		case "response.output_item.done":
+			if data.Get("item.type").String() == "function_call" {
+				doneIndexes = append(doneIndexes, data.Get("output_index").Int())
+			}
+		case "response.completed":
+			for _, item := range data.Get("response.output").Array() {
+				if item.Get("type").String() == "function_call" {
+					completedOrder = append(completedOrder, item.Get("call_id").String())
+				}
+			}
+		}
+	}
+
+	if len(doneIndexes) != 2 {
+		t.Fatalf("expected 2 function_call done indexes, got %d", len(doneIndexes))
+	}
+	if doneIndexes[0] >= doneIndexes[1] {
+		t.Fatalf("expected ascending done output indexes, got %v", doneIndexes)
+	}
+	if len(completedOrder) != 2 {
+		t.Fatalf("expected 2 function_call items in completed output, got %d", len(completedOrder))
+	}
+	if completedOrder[0] != "call_glob" || completedOrder[1] != "call_read" {
+		t.Fatalf("unexpected completed function_call order: %v", completedOrder)
+	}
+}

internal/tui/i18n.go

@@ -201,6 +201,7 @@ var zhStrings = map[string]string{
 	"usage_output":        "输出",
 	"usage_cached":        "缓存",
 	"usage_reasoning":     "思考",
+	"usage_time":          "时间",
 
 	// ── Logs ──
 	"logs_title":       "📋 日志",
@@ -352,6 +353,7 @@ var enStrings = map[string]string{
 	"usage_output":        "Output",
 	"usage_cached":        "Cached",
 	"usage_reasoning":     "Reasoning",
+	"usage_time":          "Time",
 
 	// ── Logs ──
 	"logs_title":       "📋 Logs",

internal/tui/usage_tab.go

@@ -248,6 +248,9 @@ func (m usageTabModel) renderContent() string {
 
 							// Token type breakdown from details
 							sb.WriteString(m.renderTokenBreakdown(stats))
+
+							// Latency breakdown from details
+							sb.WriteString(m.renderLatencyBreakdown(stats))
 						}
 					}
 				}
@@ -308,6 +311,57 @@ func (m usageTabModel) renderTokenBreakdown(modelStats map[string]any) string {
 		lipgloss.NewStyle().Foreground(colorMuted).Render(strings.Join(parts, "  ")))
 }
 
+// renderLatencyBreakdown aggregates latency_ms from model details and displays avg/min/max.
+func (m usageTabModel) renderLatencyBreakdown(modelStats map[string]any) string {
+	details, ok := modelStats["details"]
+	if !ok {
+		return ""
+	}
+	detailList, ok := details.([]any)
+	if !ok || len(detailList) == 0 {
+		return ""
+	}
+
+	var totalLatency int64
+	var count int
+	var minLatency, maxLatency int64
+	first := true
+
+	for _, d := range detailList {
+		dm, ok := d.(map[string]any)
+		if !ok {
+			continue
+		}
+		latencyMs := int64(getFloat(dm, "latency_ms"))
+		if latencyMs <= 0 {
+			continue
+		}
+		totalLatency += latencyMs
+		count++
+		if first {
+			minLatency = latencyMs
+			maxLatency = latencyMs
+			first = false
+		} else {
+			if latencyMs < minLatency {
+				minLatency = latencyMs
+			}
+			if latencyMs > maxLatency {
+				maxLatency = latencyMs
+			}
+		}
+	}
+
+	if count == 0 {
+		return ""
+	}
+
+	avgLatency := totalLatency / int64(count)
+	return fmt.Sprintf("    │  %s: avg %dms  min %dms  max %dms\n",
+		lipgloss.NewStyle().Foreground(colorMuted).Render(T("usage_time")),
+		avgLatency, minLatency, maxLatency)
+}
+
 // renderBarChart renders a simple ASCII horizontal bar chart.
 func renderBarChart(data map[string]any, maxBarWidth int, barColor lipgloss.Color) string {
 	if maxBarWidth < 10 {

internal/tui/usage_tab_test.go

@@ -0,0 +1,134 @@
+package tui
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestRenderLatencyBreakdown(t *testing.T) {
+	tests := []struct {
+		name         string
+		modelStats   map[string]any
+		wantEmpty    bool
+		wantContains string
+	}{
+		{
+			name:       "no details",
+			modelStats: map[string]any{},
+			wantEmpty:  true,
+		},
+		{
+			name: "empty details",
+			modelStats: map[string]any{
+				"details": []any{},
+			},
+			wantEmpty: true,
+		},
+		{
+			name: "details with zero latency",
+			modelStats: map[string]any{
+				"details": []any{
+					map[string]any{
+						"latency_ms": float64(0),
+					},
+				},
+			},
+			wantEmpty: true,
+		},
+		{
+			name: "single request with latency",
+			modelStats: map[string]any{
+				"details": []any{
+					map[string]any{
+						"latency_ms": float64(1500),
+					},
+				},
+			},
+			wantEmpty:    false,
+			wantContains: "avg 1500ms  min 1500ms  max 1500ms",
+		},
+		{
+			name: "multiple requests with varying latency",
+			modelStats: map[string]any{
+				"details": []any{
+					map[string]any{
+						"latency_ms": float64(100),
+					},
+					map[string]any{
+						"latency_ms": float64(200),
+					},
+					map[string]any{
+						"latency_ms": float64(300),
+					},
+				},
+			},
+			wantEmpty:    false,
+			wantContains: "avg 200ms  min 100ms  max 300ms",
+		},
+		{
+			name: "mixed valid and invalid latency values",
+			modelStats: map[string]any{
+				"details": []any{
+					map[string]any{
+						"latency_ms": float64(500),
+					},
+					map[string]any{
+						"latency_ms": float64(0),
+					},
+					map[string]any{
+						"latency_ms": float64(1500),
+					},
+				},
+			},
+			wantEmpty:    false,
+			wantContains: "avg 1000ms  min 500ms  max 1500ms",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			m := usageTabModel{}
+			result := m.renderLatencyBreakdown(tt.modelStats)
+
+			if tt.wantEmpty {
+				if result != "" {
+					t.Errorf("renderLatencyBreakdown() = %q, want empty string", result)
+				}
+				return
+			}
+
+			if result == "" {
+				t.Errorf("renderLatencyBreakdown() = empty, want non-empty string")
+				return
+			}
+
+			if tt.wantContains != "" && !strings.Contains(result, tt.wantContains) {
+				t.Errorf("renderLatencyBreakdown() = %q, want to contain %q", result, tt.wantContains)
+			}
+		})
+	}
+}
+
+func TestUsageTimeTranslations(t *testing.T) {
+	prevLocale := CurrentLocale()
+	t.Cleanup(func() {
+		SetLocale(prevLocale)
+	})
+
+	tests := []struct {
+		locale string
+		want   string
+	}{
+		{locale: "en", want: "Time"},
+		{locale: "zh", want: "时间"},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.locale, func(t *testing.T) {
+			SetLocale(tt.locale)
+			if got := T("usage_time"); got != tt.want {
+				t.Fatalf("T(usage_time) = %q, want %q", got, tt.want)
+			}
+		})
+	}
+}

internal/watcher/diff/config_diff.go

@@ -80,6 +80,9 @@ func BuildConfigChangeDetails(oldCfg, newCfg *config.Config) []string {
 	if oldCfg.QuotaExceeded.SwitchPreviewModel != newCfg.QuotaExceeded.SwitchPreviewModel {
 		changes = append(changes, fmt.Sprintf("quota-exceeded.switch-preview-model: %t -> %t", oldCfg.QuotaExceeded.SwitchPreviewModel, newCfg.QuotaExceeded.SwitchPreviewModel))
 	}
+	if oldCfg.QuotaExceeded.AntigravityCredits != newCfg.QuotaExceeded.AntigravityCredits {
+		changes = append(changes, fmt.Sprintf("quota-exceeded.antigravity-credits: %t -> %t", oldCfg.QuotaExceeded.AntigravityCredits, newCfg.QuotaExceeded.AntigravityCredits))
+	}
 
 	if oldCfg.Routing.Strategy != newCfg.Routing.Strategy {
 		changes = append(changes, fmt.Sprintf("routing.strategy: %s -> %s", oldCfg.Routing.Strategy, newCfg.Routing.Strategy))

internal/watcher/diff/config_diff_test.go

@@ -229,7 +229,7 @@ func TestBuildConfigChangeDetails_FlagsAndKeys(t *testing.T) {
 		MaxRetryCredentials:    1,
 		MaxRetryInterval:       1,
 		WebsocketAuth:          false,
-		QuotaExceeded:          config.QuotaExceeded{SwitchProject: false, SwitchPreviewModel: false},
+		QuotaExceeded:          config.QuotaExceeded{SwitchProject: false, SwitchPreviewModel: false, AntigravityCredits: false},
 		ClaudeKey:              []config.ClaudeKey{{APIKey: "c1"}},
 		CodexKey:               []config.CodexKey{{APIKey: "x1"}},
 		AmpCode:                config.AmpCode{UpstreamAPIKey: "keep", RestrictManagementToLocalhost: false},
@@ -253,7 +253,7 @@ func TestBuildConfigChangeDetails_FlagsAndKeys(t *testing.T) {
 		MaxRetryCredentials:    3,
 		MaxRetryInterval:       3,
 		WebsocketAuth:          true,
-		QuotaExceeded:          config.QuotaExceeded{SwitchProject: true, SwitchPreviewModel: true},
+		QuotaExceeded:          config.QuotaExceeded{SwitchProject: true, SwitchPreviewModel: true, AntigravityCredits: true},
 		ClaudeKey: []config.ClaudeKey{
 			{APIKey: "c1", BaseURL: "http://new", ProxyURL: "http://p", Headers: map[string]string{"H": "1"}, ExcludedModels: []string{"a"}},
 			{APIKey: "c2"},
@@ -297,6 +297,7 @@ func TestBuildConfigChangeDetails_FlagsAndKeys(t *testing.T) {
 	expectContains(t, details, "nonstream-keepalive-interval: 0 -> 5")
 	expectContains(t, details, "quota-exceeded.switch-project: false -> true")
 	expectContains(t, details, "quota-exceeded.switch-preview-model: false -> true")
+	expectContains(t, details, "quota-exceeded.antigravity-credits: false -> true")
 	expectContains(t, details, "api-keys count: 1 -> 2")
 	expectContains(t, details, "claude-api-key count: 1 -> 2")
 	expectContains(t, details, "codex-api-key count: 1 -> 2")
@@ -320,7 +321,7 @@ func TestBuildConfigChangeDetails_AllBranches(t *testing.T) {
 		MaxRetryCredentials:    1,
 		MaxRetryInterval:       1,
 		WebsocketAuth:          false,
-		QuotaExceeded:          config.QuotaExceeded{SwitchProject: false, SwitchPreviewModel: false},
+		QuotaExceeded:          config.QuotaExceeded{SwitchProject: false, SwitchPreviewModel: false, AntigravityCredits: false},
 		GeminiKey: []config.GeminiKey{
 			{APIKey: "g-old", BaseURL: "http://g-old", ProxyURL: "http://gp-old", Headers: map[string]string{"A": "1"}},
 		},
@@ -374,7 +375,7 @@ func TestBuildConfigChangeDetails_AllBranches(t *testing.T) {
 		MaxRetryCredentials:    3,
 		MaxRetryInterval:       3,
 		WebsocketAuth:          true,
-		QuotaExceeded:          config.QuotaExceeded{SwitchProject: true, SwitchPreviewModel: true},
+		QuotaExceeded:          config.QuotaExceeded{SwitchProject: true, SwitchPreviewModel: true, AntigravityCredits: true},
 		GeminiKey: []config.GeminiKey{
 			{APIKey: "g-new", BaseURL: "http://g-new", ProxyURL: "http://gp-new", Headers: map[string]string{"A": "2"}, ExcludedModels: []string{"x", "y"}},
 		},
@@ -437,6 +438,7 @@ func TestBuildConfigChangeDetails_AllBranches(t *testing.T) {
 	expectContains(t, changes, "ws-auth: false -> true")
 	expectContains(t, changes, "quota-exceeded.switch-project: false -> true")
 	expectContains(t, changes, "quota-exceeded.switch-preview-model: false -> true")
+	expectContains(t, changes, "quota-exceeded.antigravity-credits: false -> true")
 	expectContains(t, changes, "api-keys: values updated (count unchanged, redacted)")
 	expectContains(t, changes, "gemini[0].base-url: http://g-old -> http://g-new")
 	expectContains(t, changes, "gemini[0].proxy-url: http://gp-old -> http://gp-new")

internal/watcher/synthesizer/file.go

@@ -157,6 +157,7 @@ func synthesizeFileAuths(ctx *SynthesisContext, fullPath string, data []byte) []
 			}
 		}
 	}
+	coreauth.ApplyCustomHeadersFromMetadata(a)
 	ApplyAuthExcludedModelsMeta(a, cfg, perAccountExcluded, "oauth")
 	// For codex auth files, extract plan_type from the JWT id_token.
 	if provider == "codex" {
@@ -233,6 +234,11 @@ func SynthesizeGeminiVirtualAuths(primary *coreauth.Auth, metadata map[string]an
 		if noteVal, hasNote := primary.Attributes["note"]; hasNote && noteVal != "" {
 			attrs["note"] = noteVal
 		}
+		for k, v := range primary.Attributes {
+			if strings.HasPrefix(k, "header:") && strings.TrimSpace(v) != "" {
+				attrs[k] = v
+			}
+		}
 		metadataCopy := map[string]any{
 			"email":             email,
 			"project_id":        projectID,

internal/watcher/synthesizer/file_test.go

@@ -69,10 +69,14 @@ func TestFileSynthesizer_Synthesize_ValidAuthFile(t *testing.T) {
 
 	// Create a valid auth file
 	authData := map[string]any{
-		"type":            "claude",
-		"email":           "test@example.com",
-		"proxy_url":       "http://proxy.local",
-		"prefix":          "test-prefix",
+		"type":      "claude",
+		"email":     "test@example.com",
+		"proxy_url": "http://proxy.local",
+		"prefix":    "test-prefix",
+		"headers": map[string]string{
+			" X-Test ": " value ",
+			"X-Empty":  "  ",
+		},
 		"disable_cooling": true,
 		"request_retry":   2,
 	}
@@ -110,6 +114,12 @@ func TestFileSynthesizer_Synthesize_ValidAuthFile(t *testing.T) {
 	if auths[0].ProxyURL != "http://proxy.local" {
 		t.Errorf("expected proxy_url http://proxy.local, got %s", auths[0].ProxyURL)
 	}
+	if got := auths[0].Attributes["header:X-Test"]; got != "value" {
+		t.Errorf("expected header:X-Test value, got %q", got)
+	}
+	if _, ok := auths[0].Attributes["header:X-Empty"]; ok {
+		t.Errorf("expected header:X-Empty to be absent, got %q", auths[0].Attributes["header:X-Empty"])
+	}
 	if v, ok := auths[0].Metadata["disable_cooling"].(bool); !ok || !v {
 		t.Errorf("expected disable_cooling true, got %v", auths[0].Metadata["disable_cooling"])
 	}
@@ -450,8 +460,9 @@ func TestSynthesizeGeminiVirtualAuths_MultiProject(t *testing.T) {
 		Prefix:   "test-prefix",
 		ProxyURL: "http://proxy.local",
 		Attributes: map[string]string{
-			"source": "test-source",
-			"path":   "/path/to/auth",
+			"source":       "test-source",
+			"path":         "/path/to/auth",
+			"header:X-Tra": "value",
 		},
 	}
 	metadata := map[string]any{
@@ -506,6 +517,9 @@ func TestSynthesizeGeminiVirtualAuths_MultiProject(t *testing.T) {
 		if v.Attributes["runtime_only"] != "true" {
 			t.Error("expected runtime_only=true")
 		}
+		if got := v.Attributes["header:X-Tra"]; got != "value" {
+			t.Errorf("expected virtual %d header:X-Tra %q, got %q", i, "value", got)
+		}
 		if v.Attributes["gemini_virtual_parent"] != "primary-id" {
 			t.Errorf("expected gemini_virtual_parent=primary-id, got %s", v.Attributes["gemini_virtual_parent"])
 		}

sdk/api/handlers/gemini/gemini-cli_handlers.go

@@ -9,6 +9,7 @@ import (
 	"context"
 	"fmt"
 	"io"
+	"net"
 	"net/http"
 	"strings"
 	"time"
@@ -49,7 +50,23 @@ func (h *GeminiCLIAPIHandler) Models() []map[string]any {
 // CLIHandler handles CLI-specific requests for Gemini API operations.
 // It restricts access to localhost only and routes requests to appropriate internal handlers.
 func (h *GeminiCLIAPIHandler) CLIHandler(c *gin.Context) {
-	if !strings.HasPrefix(c.Request.RemoteAddr, "127.0.0.1:") {
+	if h.Cfg == nil || !h.Cfg.EnableGeminiCLIEndpoint {
+		c.JSON(http.StatusForbidden, handlers.ErrorResponse{
+			Error: handlers.ErrorDetail{
+				Message: "Gemini CLI endpoint is disabled",
+				Type:    "forbidden",
+			},
+		})
+		return
+	}
+
+	requestHost := c.Request.Host
+	requestHostname := requestHost
+	if hostname, _, errSplitHostPort := net.SplitHostPort(requestHost); errSplitHostPort == nil {
+		requestHostname = hostname
+	}
+
+	if !strings.HasPrefix(c.Request.RemoteAddr, "127.0.0.1:") || requestHostname != "127.0.0.1" {
 		c.JSON(http.StatusForbidden, handlers.ErrorResponse{
 			Error: handlers.ErrorDetail{
 				Message: "CLI reply only allow local access",

sdk/api/handlers/handlers_stream_bootstrap_test.go

@@ -136,6 +136,8 @@ type authAwareStreamExecutor struct {
 
 type invalidJSONStreamExecutor struct{}
 
+type splitResponsesEventStreamExecutor struct{}
+
 func (e *invalidJSONStreamExecutor) Identifier() string { return "codex" }
 
 func (e *invalidJSONStreamExecutor) Execute(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (coreexecutor.Response, error) {
@@ -165,6 +167,36 @@ func (e *invalidJSONStreamExecutor) HttpRequest(ctx context.Context, auth *corea
 	}
 }
 
+func (e *splitResponsesEventStreamExecutor) Identifier() string { return "split-sse" }
+
+func (e *splitResponsesEventStreamExecutor) Execute(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (coreexecutor.Response, error) {
+	return coreexecutor.Response{}, &coreauth.Error{Code: "not_implemented", Message: "Execute not implemented"}
+}
+
+func (e *splitResponsesEventStreamExecutor) ExecuteStream(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (*coreexecutor.StreamResult, error) {
+	ch := make(chan coreexecutor.StreamChunk, 2)
+	ch <- coreexecutor.StreamChunk{Payload: []byte("event: response.completed")}
+	ch <- coreexecutor.StreamChunk{Payload: []byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp-1\",\"output\":[]}}")}
+	close(ch)
+	return &coreexecutor.StreamResult{Chunks: ch}, nil
+}
+
+func (e *splitResponsesEventStreamExecutor) Refresh(ctx context.Context, auth *coreauth.Auth) (*coreauth.Auth, error) {
+	return auth, nil
+}
+
+func (e *splitResponsesEventStreamExecutor) CountTokens(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (coreexecutor.Response, error) {
+	return coreexecutor.Response{}, &coreauth.Error{Code: "not_implemented", Message: "CountTokens not implemented"}
+}
+
+func (e *splitResponsesEventStreamExecutor) HttpRequest(ctx context.Context, auth *coreauth.Auth, req *http.Request) (*http.Response, error) {
+	return nil, &coreauth.Error{
+		Code:       "not_implemented",
+		Message:    "HttpRequest not implemented",
+		HTTPStatus: http.StatusNotImplemented,
+	}
+}
+
 func (e *authAwareStreamExecutor) Identifier() string { return "codex" }
 
 func (e *authAwareStreamExecutor) Execute(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (coreexecutor.Response, error) {
@@ -607,3 +639,52 @@ func TestExecuteStreamWithAuthManager_ValidatesOpenAIResponsesStreamDataJSON(t *
 		t.Fatalf("expected terminal error")
 	}
 }
+
+func TestExecuteStreamWithAuthManager_AllowsSplitOpenAIResponsesSSEEventLines(t *testing.T) {
+	executor := &splitResponsesEventStreamExecutor{}
+	manager := coreauth.NewManager(nil, nil, nil)
+	manager.RegisterExecutor(executor)
+
+	auth1 := &coreauth.Auth{
+		ID:       "auth1",
+		Provider: "split-sse",
+		Status:   coreauth.StatusActive,
+		Metadata: map[string]any{"email": "test1@example.com"},
+	}
+	if _, err := manager.Register(context.Background(), auth1); err != nil {
+		t.Fatalf("manager.Register(auth1): %v", err)
+	}
+
+	registry.GetGlobalRegistry().RegisterClient(auth1.ID, auth1.Provider, []*registry.ModelInfo{{ID: "test-model"}})
+	t.Cleanup(func() {
+		registry.GetGlobalRegistry().UnregisterClient(auth1.ID)
+	})
+
+	handler := NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, manager)
+	dataChan, _, errChan := handler.ExecuteStreamWithAuthManager(context.Background(), "openai-response", "test-model", []byte(`{"model":"test-model"}`), "")
+	if dataChan == nil || errChan == nil {
+		t.Fatalf("expected non-nil channels")
+	}
+
+	var got []string
+	for chunk := range dataChan {
+		got = append(got, string(chunk))
+	}
+
+	for msg := range errChan {
+		if msg != nil {
+			t.Fatalf("unexpected error: %+v", msg)
+		}
+	}
+
+	if len(got) != 2 {
+		t.Fatalf("expected 2 forwarded chunks, got %d: %#v", len(got), got)
+	}
+	if got[0] != "event: response.completed" {
+		t.Fatalf("unexpected first chunk: %q", got[0])
+	}
+	expectedData := "data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp-1\",\"output\":[]}}"
+	if got[1] != expectedData {
+		t.Fatalf("unexpected second chunk.\nGot:  %q\nWant: %q", got[1], expectedData)
+	}
+}

sdk/api/handlers/header_filter.go

@@ -5,6 +5,18 @@ import (
 	"strings"
 )
 
+// gatewayHeaderPrefixes lists header name prefixes injected by known AI gateway
+// proxies. Claude Code's client-side telemetry detects these and reports the
+// gateway type, so we strip them from upstream responses to avoid detection.
+var gatewayHeaderPrefixes = []string{
+	"x-litellm-",
+	"helicone-",
+	"x-portkey-",
+	"cf-aig-",
+	"x-kong-",
+	"x-bt-",
+}
+
 // hopByHopHeaders lists RFC 7230 Section 6.1 hop-by-hop headers that MUST NOT
 // be forwarded by proxies, plus security-sensitive headers that should not leak.
 var hopByHopHeaders = map[string]struct{}{
@@ -40,6 +52,19 @@ func FilterUpstreamHeaders(src http.Header) http.Header {
 		if _, scoped := connectionScoped[canonicalKey]; scoped {
 			continue
 		}
+		// Strip headers injected by known AI gateway proxies to avoid
+		// Claude Code client-side gateway detection.
+		lowerKey := strings.ToLower(key)
+		gatewayMatch := false
+		for _, prefix := range gatewayHeaderPrefixes {
+			if strings.HasPrefix(lowerKey, prefix) {
+				gatewayMatch = true
+				break
+			}
+		}
+		if gatewayMatch {
+			continue
+		}
 		dst[key] = values
 	}
 	if len(dst) == 0 {

sdk/api/handlers/openai/endpoint_compat.go

@@ -1,6 +1,9 @@
 package openai
 
-import "github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+import (
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/thinking"
+)
 
 const (
 	openAIChatEndpoint      = "/chat/completions"
@@ -12,6 +15,12 @@ func resolveEndpointOverride(modelName, requestedEndpoint string) (string, bool)
 		return "", false
 	}
 	info := registry.GetGlobalRegistry().GetModelInfo(modelName, "")
+	if info == nil {
+		baseModel := thinking.ParseSuffix(modelName).ModelName
+		if baseModel != "" && baseModel != modelName {
+			info = registry.GetGlobalRegistry().GetModelInfo(baseModel, "")
+		}
+	}
 	if info == nil || len(info.SupportedEndpoints) == 0 {
 		return "", false
 	}

sdk/api/handlers/openai/endpoint_compat_test.go

@@ -0,0 +1,29 @@
+package openai
+
+import (
+	"testing"
+
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+)
+
+func TestResolveEndpointOverride_StripsThinkingSuffix(t *testing.T) {
+	const clientID = "test-endpoint-compat-suffix"
+	reg := registry.GetGlobalRegistry()
+	reg.RegisterClient(clientID, "github-copilot", []*registry.ModelInfo{
+		{
+			ID:                 "test-gemini-chat-only",
+			SupportedEndpoints: []string{openAIChatEndpoint},
+		},
+	})
+	t.Cleanup(func() {
+		reg.UnregisterClient(clientID)
+	})
+
+	override, ok := resolveEndpointOverride("test-gemini-chat-only(high)", openAIResponsesEndpoint)
+	if !ok {
+		t.Fatalf("expected endpoint override to be resolved")
+	}
+	if override != openAIChatEndpoint {
+		t.Fatalf("override endpoint = %q, want %q", override, openAIChatEndpoint)
+	}
+}

sdk/api/handlers/openai/openai_responses_handlers.go

@@ -9,7 +9,9 @@ package openai
 import (
 	"bytes"
 	"context"
+	"encoding/json"
 	"fmt"
+	"io"
 	"net/http"
 
 	"github.com/gin-gonic/gin"
@@ -22,6 +24,177 @@ import (
 	"github.com/tidwall/sjson"
 )
 
+func writeResponsesSSEChunk(w io.Writer, chunk []byte) {
+	if w == nil || len(chunk) == 0 {
+		return
+	}
+	if _, err := w.Write(chunk); err != nil {
+		return
+	}
+	if bytes.HasSuffix(chunk, []byte("\n\n")) || bytes.HasSuffix(chunk, []byte("\r\n\r\n")) {
+		return
+	}
+	suffix := []byte("\n\n")
+	if bytes.HasSuffix(chunk, []byte("\r\n")) {
+		suffix = []byte("\r\n")
+	} else if bytes.HasSuffix(chunk, []byte("\n")) {
+		suffix = []byte("\n")
+	}
+	if _, err := w.Write(suffix); err != nil {
+		return
+	}
+}
+
+type responsesSSEFramer struct {
+	pending []byte
+}
+
+func (f *responsesSSEFramer) WriteChunk(w io.Writer, chunk []byte) {
+	if len(chunk) == 0 {
+		return
+	}
+	if responsesSSENeedsLineBreak(f.pending, chunk) {
+		f.pending = append(f.pending, '\n')
+	}
+	f.pending = append(f.pending, chunk...)
+	for {
+		frameLen := responsesSSEFrameLen(f.pending)
+		if frameLen == 0 {
+			break
+		}
+		writeResponsesSSEChunk(w, f.pending[:frameLen])
+		copy(f.pending, f.pending[frameLen:])
+		f.pending = f.pending[:len(f.pending)-frameLen]
+	}
+	if len(bytes.TrimSpace(f.pending)) == 0 {
+		f.pending = f.pending[:0]
+		return
+	}
+	if len(f.pending) == 0 || !responsesSSECanEmitWithoutDelimiter(f.pending) {
+		return
+	}
+	writeResponsesSSEChunk(w, f.pending)
+	f.pending = f.pending[:0]
+}
+
+func (f *responsesSSEFramer) Flush(w io.Writer) {
+	if len(f.pending) == 0 {
+		return
+	}
+	if len(bytes.TrimSpace(f.pending)) == 0 {
+		f.pending = f.pending[:0]
+		return
+	}
+	if !responsesSSECanEmitWithoutDelimiter(f.pending) {
+		f.pending = f.pending[:0]
+		return
+	}
+	writeResponsesSSEChunk(w, f.pending)
+	f.pending = f.pending[:0]
+}
+
+func responsesSSEFrameLen(chunk []byte) int {
+	if len(chunk) == 0 {
+		return 0
+	}
+	lf := bytes.Index(chunk, []byte("\n\n"))
+	crlf := bytes.Index(chunk, []byte("\r\n\r\n"))
+	switch {
+	case lf < 0:
+		if crlf < 0 {
+			return 0
+		}
+		return crlf + 4
+	case crlf < 0:
+		return lf + 2
+	case lf < crlf:
+		return lf + 2
+	default:
+		return crlf + 4
+	}
+}
+
+func responsesSSENeedsMoreData(chunk []byte) bool {
+	trimmed := bytes.TrimSpace(chunk)
+	if len(trimmed) == 0 {
+		return false
+	}
+	return responsesSSEHasField(trimmed, []byte("event:")) && !responsesSSEHasField(trimmed, []byte("data:"))
+}
+
+func responsesSSEHasField(chunk []byte, prefix []byte) bool {
+	s := chunk
+	for len(s) > 0 {
+		line := s
+		if i := bytes.IndexByte(s, '\n'); i >= 0 {
+			line = s[:i]
+			s = s[i+1:]
+		} else {
+			s = nil
+		}
+		line = bytes.TrimSpace(line)
+		if bytes.HasPrefix(line, prefix) {
+			return true
+		}
+	}
+	return false
+}
+
+func responsesSSECanEmitWithoutDelimiter(chunk []byte) bool {
+	trimmed := bytes.TrimSpace(chunk)
+	if len(trimmed) == 0 || responsesSSENeedsMoreData(trimmed) || !responsesSSEHasField(trimmed, []byte("data:")) {
+		return false
+	}
+	return responsesSSEDataLinesValid(trimmed)
+}
+
+func responsesSSEDataLinesValid(chunk []byte) bool {
+	s := chunk
+	for len(s) > 0 {
+		line := s
+		if i := bytes.IndexByte(s, '\n'); i >= 0 {
+			line = s[:i]
+			s = s[i+1:]
+		} else {
+			s = nil
+		}
+		line = bytes.TrimSpace(line)
+		if len(line) == 0 || !bytes.HasPrefix(line, []byte("data:")) {
+			continue
+		}
+		data := bytes.TrimSpace(line[len("data:"):])
+		if len(data) == 0 || bytes.Equal(data, []byte("[DONE]")) {
+			continue
+		}
+		if !json.Valid(data) {
+			return false
+		}
+	}
+	return true
+}
+
+func responsesSSENeedsLineBreak(pending, chunk []byte) bool {
+	if len(pending) == 0 || len(chunk) == 0 {
+		return false
+	}
+	if bytes.HasSuffix(pending, []byte("\n")) || bytes.HasSuffix(pending, []byte("\r")) {
+		return false
+	}
+	if chunk[0] == '\n' || chunk[0] == '\r' {
+		return false
+	}
+	trimmed := bytes.TrimLeft(chunk, " \t")
+	if len(trimmed) == 0 {
+		return false
+	}
+	for _, prefix := range [][]byte{[]byte("data:"), []byte("event:"), []byte("id:"), []byte("retry:"), []byte(":")} {
+		if bytes.HasPrefix(trimmed, prefix) {
+			return true
+		}
+	}
+	return false
+}
+
 // OpenAIResponsesAPIHandler contains the handlers for OpenAIResponses API endpoints.
 // It holds a pool of clients to interact with the backend service.
 type OpenAIResponsesAPIHandler struct {
@@ -234,6 +407,7 @@ func (h *OpenAIResponsesAPIHandler) handleStreamingResponse(c *gin.Context, rawJ
 		c.Header("Connection", "keep-alive")
 		c.Header("Access-Control-Allow-Origin", "*")
 	}
+	framer := &responsesSSEFramer{}
 
 	// Peek at the first chunk
 	for {
@@ -271,15 +445,11 @@ func (h *OpenAIResponsesAPIHandler) handleStreamingResponse(c *gin.Context, rawJ
 			handlers.WriteUpstreamHeaders(c.Writer.Header(), upstreamHeaders)
 
 			// Write first chunk logic (matching forwardResponsesStream)
-			if bytes.HasPrefix(chunk, []byte("event:")) {
-				_, _ = c.Writer.Write([]byte("\n"))
-			}
-			_, _ = c.Writer.Write(chunk)
-			_, _ = c.Writer.Write([]byte("\n"))
+			framer.WriteChunk(c.Writer, chunk)
 			flusher.Flush()
 
 			// Continue
-			h.forwardResponsesStream(c, flusher, func(err error) { cliCancel(err) }, dataChan, errChan)
+			h.forwardResponsesStream(c, flusher, func(err error) { cliCancel(err) }, dataChan, errChan, framer)
 			return
 		}
 	}
@@ -397,16 +567,16 @@ func (h *OpenAIResponsesAPIHandler) forwardChatAsResponsesStream(c *gin.Context,
 	})
 }
 
-func (h *OpenAIResponsesAPIHandler) forwardResponsesStream(c *gin.Context, flusher http.Flusher, cancel func(error), data <-chan []byte, errs <-chan *interfaces.ErrorMessage) {
+func (h *OpenAIResponsesAPIHandler) forwardResponsesStream(c *gin.Context, flusher http.Flusher, cancel func(error), data <-chan []byte, errs <-chan *interfaces.ErrorMessage, framer *responsesSSEFramer) {
+	if framer == nil {
+		framer = &responsesSSEFramer{}
+	}
 	h.ForwardStream(c, flusher, cancel, data, errs, handlers.StreamForwardOptions{
 		WriteChunk: func(chunk []byte) {
-			if bytes.HasPrefix(chunk, []byte("event:")) {
-				_, _ = c.Writer.Write([]byte("\n"))
-			}
-			_, _ = c.Writer.Write(chunk)
-			_, _ = c.Writer.Write([]byte("\n"))
+			framer.WriteChunk(c.Writer, chunk)
 		},
 		WriteTerminalError: func(errMsg *interfaces.ErrorMessage) {
+			framer.Flush(c.Writer)
 			if errMsg == nil {
 				return
 			}
@@ -422,6 +592,7 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesStream(c *gin.Context, flush
 			_, _ = fmt.Fprintf(c.Writer, "\nevent: error\ndata: %s\n\n", string(chunk))
 		},
 		WriteDone: func() {
+			framer.Flush(c.Writer)
 			_, _ = c.Writer.Write([]byte("\n"))
 		},
 	})

sdk/api/handlers/openai/openai_responses_handlers_stream_error_test.go

@@ -32,7 +32,7 @@ func TestForwardResponsesStreamTerminalErrorUsesResponsesErrorChunk(t *testing.T
 	errs <- &interfaces.ErrorMessage{StatusCode: http.StatusInternalServerError, Error: errors.New("unexpected EOF")}
 	close(errs)
 
-	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs)
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
 	body := recorder.Body.String()
 	if !strings.Contains(body, `"type":"error"`) {
 		t.Fatalf("expected responses error chunk, got: %q", body)

sdk/api/handlers/openai/openai_responses_handlers_stream_test.go

@@ -0,0 +1,142 @@
+package openai
+
+import (
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"github.com/gin-gonic/gin"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/interfaces"
+	"github.com/router-for-me/CLIProxyAPI/v6/sdk/api/handlers"
+	sdkconfig "github.com/router-for-me/CLIProxyAPI/v6/sdk/config"
+)
+
+func newResponsesStreamTestHandler(t *testing.T) (*OpenAIResponsesAPIHandler, *httptest.ResponseRecorder, *gin.Context, http.Flusher) {
+	t.Helper()
+
+	gin.SetMode(gin.TestMode)
+	base := handlers.NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, nil)
+	h := NewOpenAIResponsesAPIHandler(base)
+
+	recorder := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(recorder)
+	c.Request = httptest.NewRequest(http.MethodPost, "/v1/responses", nil)
+
+	flusher, ok := c.Writer.(http.Flusher)
+	if !ok {
+		t.Fatalf("expected gin writer to implement http.Flusher")
+	}
+
+	return h, recorder, c, flusher
+}
+
+func TestForwardResponsesStreamSeparatesDataOnlySSEChunks(t *testing.T) {
+	h, recorder, c, flusher := newResponsesStreamTestHandler(t)
+
+	data := make(chan []byte, 2)
+	errs := make(chan *interfaces.ErrorMessage)
+	data <- []byte("data: {\"type\":\"response.output_item.done\",\"item\":{\"type\":\"function_call\",\"arguments\":\"{}\"}}")
+	data <- []byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp-1\",\"output\":[]}}")
+	close(data)
+	close(errs)
+
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
+	body := recorder.Body.String()
+	parts := strings.Split(strings.TrimSpace(body), "\n\n")
+	if len(parts) != 2 {
+		t.Fatalf("expected 2 SSE events, got %d. Body: %q", len(parts), body)
+	}
+
+	expectedPart1 := "data: {\"type\":\"response.output_item.done\",\"item\":{\"type\":\"function_call\",\"arguments\":\"{}\"}}"
+	if parts[0] != expectedPart1 {
+		t.Errorf("unexpected first event.\nGot: %q\nWant: %q", parts[0], expectedPart1)
+	}
+
+	expectedPart2 := "data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp-1\",\"output\":[]}}"
+	if parts[1] != expectedPart2 {
+		t.Errorf("unexpected second event.\nGot: %q\nWant: %q", parts[1], expectedPart2)
+	}
+}
+
+func TestForwardResponsesStreamReassemblesSplitSSEEventChunks(t *testing.T) {
+	h, recorder, c, flusher := newResponsesStreamTestHandler(t)
+
+	data := make(chan []byte, 3)
+	errs := make(chan *interfaces.ErrorMessage)
+	data <- []byte("event: response.created")
+	data <- []byte("data: {\"type\":\"response.created\",\"response\":{\"id\":\"resp-1\"}}")
+	data <- []byte("\n")
+	close(data)
+	close(errs)
+
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
+
+	got := strings.TrimSuffix(recorder.Body.String(), "\n")
+	want := "event: response.created\ndata: {\"type\":\"response.created\",\"response\":{\"id\":\"resp-1\"}}\n\n"
+	if got != want {
+		t.Fatalf("unexpected split-event framing.\nGot:  %q\nWant: %q", got, want)
+	}
+}
+
+func TestForwardResponsesStreamPreservesValidFullSSEEventChunks(t *testing.T) {
+	h, recorder, c, flusher := newResponsesStreamTestHandler(t)
+
+	data := make(chan []byte, 1)
+	errs := make(chan *interfaces.ErrorMessage)
+	chunk := []byte("event: response.created\ndata: {\"type\":\"response.created\",\"response\":{\"id\":\"resp-1\"}}\n\n")
+	data <- chunk
+	close(data)
+	close(errs)
+
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
+
+	got := strings.TrimSuffix(recorder.Body.String(), "\n")
+	if got != string(chunk) {
+		t.Fatalf("unexpected full-event framing.\nGot:  %q\nWant: %q", got, string(chunk))
+	}
+}
+
+func TestForwardResponsesStreamBuffersSplitDataPayloadChunks(t *testing.T) {
+	h, recorder, c, flusher := newResponsesStreamTestHandler(t)
+
+	data := make(chan []byte, 2)
+	errs := make(chan *interfaces.ErrorMessage)
+	data <- []byte("data: {\"type\":\"response.created\"")
+	data <- []byte(",\"response\":{\"id\":\"resp-1\"}}")
+	close(data)
+	close(errs)
+
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
+
+	got := recorder.Body.String()
+	want := "data: {\"type\":\"response.created\",\"response\":{\"id\":\"resp-1\"}}\n\n\n"
+	if got != want {
+		t.Fatalf("unexpected split-data framing.\nGot:  %q\nWant: %q", got, want)
+	}
+}
+
+func TestResponsesSSENeedsLineBreakSkipsChunksThatAlreadyStartWithNewline(t *testing.T) {
+	if responsesSSENeedsLineBreak([]byte("event: response.created"), []byte("\n")) {
+		t.Fatal("expected no injected newline before newline-only chunk")
+	}
+	if responsesSSENeedsLineBreak([]byte("event: response.created"), []byte("\r\n")) {
+		t.Fatal("expected no injected newline before CRLF chunk")
+	}
+}
+
+func TestForwardResponsesStreamDropsIncompleteTrailingDataChunkOnFlush(t *testing.T) {
+	h, recorder, c, flusher := newResponsesStreamTestHandler(t)
+
+	data := make(chan []byte, 1)
+	errs := make(chan *interfaces.ErrorMessage)
+	data <- []byte("data: {\"type\":\"response.created\"")
+	close(data)
+	close(errs)
+
+	h.forwardResponsesStream(c, flusher, func(error) {}, data, errs, nil)
+
+	if got := recorder.Body.String(); got != "\n" {
+		t.Fatalf("expected incomplete trailing data to be dropped on flush.\nGot: %q", got)
+	}
+}

sdk/api/handlers/openai/openai_responses_websocket.go

@@ -32,10 +32,7 @@ const (
 	wsEventTypeCompleted = "response.completed"
 	wsDoneMarker         = "[DONE]"
 	wsTurnStateHeader    = "x-codex-turn-state"
-	wsRequestBodyKey     = "REQUEST_BODY_OVERRIDE"
-	wsPayloadLogMaxSize  = 2048
-	wsBodyLogMaxSize     = 64 * 1024
-	wsBodyLogTruncated   = "\n[websocket log truncated]\n"
+	wsTimelineBodyKey    = "WEBSOCKET_TIMELINE_OVERRIDE"
 )
 
 var responsesWebsocketUpgrader = websocket.Upgrader{
@@ -55,15 +52,16 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 		return
 	}
 	passthroughSessionID := uuid.NewString()
-	clientRemoteAddr := ""
-	if c != nil && c.Request != nil {
-		clientRemoteAddr = strings.TrimSpace(c.Request.RemoteAddr)
-	}
-	log.Infof("responses websocket: client connected id=%s remote=%s", passthroughSessionID, clientRemoteAddr)
+	downstreamSessionKey := websocketDownstreamSessionKey(c.Request)
+	retainResponsesWebsocketToolCaches(downstreamSessionKey)
+	clientIP := websocketClientAddress(c)
+	log.Infof("responses websocket: client connected id=%s remote=%s", passthroughSessionID, clientIP)
 	var wsTerminateErr error
-	var wsBodyLog strings.Builder
+	var wsTimelineLog strings.Builder
 	defer func() {
+		releaseResponsesWebsocketToolCaches(downstreamSessionKey)
 		if wsTerminateErr != nil {
+			appendWebsocketTimelineDisconnect(&wsTimelineLog, wsTerminateErr, time.Now())
 			// log.Infof("responses websocket: session closing id=%s reason=%v", passthroughSessionID, wsTerminateErr)
 		} else {
 			log.Infof("responses websocket: session closing id=%s", passthroughSessionID)
@@ -72,7 +70,7 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 			h.AuthManager.CloseExecutionSession(passthroughSessionID)
 			log.Infof("responses websocket: upstream execution session closed id=%s", passthroughSessionID)
 		}
-		setWebsocketRequestBody(c, wsBodyLog.String())
+		setWebsocketTimelineBody(c, wsTimelineLog.String())
 		if errClose := conn.Close(); errClose != nil {
 			log.Warnf("responses websocket: close connection error: %v", errClose)
 		}
@@ -86,7 +84,6 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 		msgType, payload, errReadMessage := conn.ReadMessage()
 		if errReadMessage != nil {
 			wsTerminateErr = errReadMessage
-			appendWebsocketEvent(&wsBodyLog, "disconnect", []byte(errReadMessage.Error()))
 			if websocket.IsCloseError(errReadMessage, websocket.CloseNormalClosure, websocket.CloseGoingAway, websocket.CloseNoStatusReceived) {
 				log.Infof("responses websocket: client disconnected id=%s error=%v", passthroughSessionID, errReadMessage)
 			} else {
@@ -104,7 +101,7 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 		// 	websocketPayloadEventType(payload),
 		// 	websocketPayloadPreview(payload),
 		// )
-		appendWebsocketEvent(&wsBodyLog, "request", payload)
+		appendWebsocketTimelineEvent(&wsTimelineLog, "request", payload, time.Now())
 
 		allowIncrementalInputWithPreviousResponseID := false
 		if pinnedAuthID != "" && h != nil && h.AuthManager != nil {
@@ -131,8 +128,7 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 		if errMsg != nil {
 			h.LoggingAPIResponseError(context.WithValue(context.Background(), "gin", c), errMsg)
 			markAPIResponseTimestamp(c)
-			errorPayload, errWrite := writeResponsesWebsocketError(conn, errMsg)
-			appendWebsocketEvent(&wsBodyLog, "response", errorPayload)
+			errorPayload, errWrite := writeResponsesWebsocketError(conn, &wsTimelineLog, errMsg)
 			log.Infof(
 				"responses websocket: downstream_out id=%s type=%d event=%s payload=%s",
 				passthroughSessionID,
@@ -160,13 +156,15 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 			}
 			lastRequest = updatedLastRequest
 			lastResponseOutput = []byte("[]")
-			if errWrite := writeResponsesWebsocketSyntheticPrewarm(c, conn, requestJSON, &wsBodyLog, passthroughSessionID); errWrite != nil {
+			if errWrite := writeResponsesWebsocketSyntheticPrewarm(c, conn, requestJSON, &wsTimelineLog, passthroughSessionID); errWrite != nil {
 				wsTerminateErr = errWrite
-				appendWebsocketEvent(&wsBodyLog, "disconnect", []byte(errWrite.Error()))
 				return
 			}
 			continue
 		}
+
+		requestJSON = repairResponsesWebsocketToolCalls(downstreamSessionKey, requestJSON)
+		updatedLastRequest = bytes.Clone(requestJSON)
 		lastRequest = updatedLastRequest
 
 		modelName := gjson.GetBytes(requestJSON, "model").String()
@@ -192,10 +190,9 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 		}
 		dataChan, _, errChan := h.ExecuteStreamWithAuthManager(cliCtx, h.HandlerType(), modelName, requestJSON, "")
 
-		completedOutput, errForward := h.forwardResponsesWebsocket(c, conn, cliCancel, dataChan, errChan, &wsBodyLog, passthroughSessionID)
+		completedOutput, errForward := h.forwardResponsesWebsocket(c, conn, cliCancel, dataChan, errChan, &wsTimelineLog, passthroughSessionID)
 		if errForward != nil {
 			wsTerminateErr = errForward
-			appendWebsocketEvent(&wsBodyLog, "disconnect", []byte(errForward.Error()))
 			log.Warnf("responses websocket: forward failed id=%s error=%v", passthroughSessionID, errForward)
 			return
 		}
@@ -203,6 +200,13 @@ func (h *OpenAIResponsesAPIHandler) ResponsesWebsocket(c *gin.Context) {
 	}
 }
 
+func websocketClientAddress(c *gin.Context) string {
+	if c == nil || c.Request == nil {
+		return ""
+	}
+	return strings.TrimSpace(c.ClientIP())
+}
+
 func websocketUpgradeHeaders(req *http.Request) http.Header {
 	headers := http.Header{}
 	if req == nil {
@@ -277,6 +281,15 @@ func normalizeResponseSubsequentRequest(rawJSON []byte, lastRequest []byte, last
 		}
 	}
 
+	// Compaction can cause clients to replace local websocket history with a new
+	// compact transcript on the next `response.create`. When the input already
+	// contains historical model output items, treating it as an incremental append
+	// duplicates stale turn-state and can leave late orphaned function_call items.
+	if shouldReplaceWebsocketTranscript(rawJSON, nextInput) {
+		normalized := normalizeResponseTranscriptReplacement(rawJSON, lastRequest)
+		return normalized, bytes.Clone(normalized), nil
+	}
+
 	// Websocket v2 mode uses response.create with previous_response_id + incremental input.
 	// Do not expand it into a full input transcript; upstream expects the incremental payload.
 	if allowIncrementalInputWithPreviousResponseID {
@@ -318,6 +331,10 @@ func normalizeResponseSubsequentRequest(rawJSON []byte, lastRequest []byte, last
 			Error:      fmt.Errorf("invalid request input: %w", errMerge),
 		}
 	}
+	dedupedInput, errDedupeFunctionCalls := dedupeFunctionCallsByCallID(mergedInput)
+	if errDedupeFunctionCalls == nil {
+		mergedInput = dedupedInput
+	}
 
 	normalized, errDelete := sjson.DeleteBytes(rawJSON, "type")
 	if errDelete != nil {
@@ -348,6 +365,91 @@ func normalizeResponseSubsequentRequest(rawJSON []byte, lastRequest []byte, last
 	return normalized, bytes.Clone(normalized), nil
 }
 
+func shouldReplaceWebsocketTranscript(rawJSON []byte, nextInput gjson.Result) bool {
+	requestType := strings.TrimSpace(gjson.GetBytes(rawJSON, "type").String())
+	if requestType != wsRequestTypeCreate && requestType != wsRequestTypeAppend {
+		return false
+	}
+	if strings.TrimSpace(gjson.GetBytes(rawJSON, "previous_response_id").String()) != "" {
+		return false
+	}
+	if !nextInput.Exists() || !nextInput.IsArray() {
+		return false
+	}
+
+	for _, item := range nextInput.Array() {
+		switch strings.TrimSpace(item.Get("type").String()) {
+		case "function_call", "custom_tool_call":
+			return true
+		case "message":
+			role := strings.TrimSpace(item.Get("role").String())
+			if role == "assistant" {
+				return true
+			}
+		}
+	}
+
+	return false
+}
+
+func normalizeResponseTranscriptReplacement(rawJSON []byte, lastRequest []byte) []byte {
+	normalized, errDelete := sjson.DeleteBytes(rawJSON, "type")
+	if errDelete != nil {
+		normalized = bytes.Clone(rawJSON)
+	}
+	normalized, _ = sjson.DeleteBytes(normalized, "previous_response_id")
+	if !gjson.GetBytes(normalized, "model").Exists() {
+		modelName := strings.TrimSpace(gjson.GetBytes(lastRequest, "model").String())
+		if modelName != "" {
+			normalized, _ = sjson.SetBytes(normalized, "model", modelName)
+		}
+	}
+	if !gjson.GetBytes(normalized, "instructions").Exists() {
+		instructions := gjson.GetBytes(lastRequest, "instructions")
+		if instructions.Exists() {
+			normalized, _ = sjson.SetRawBytes(normalized, "instructions", []byte(instructions.Raw))
+		}
+	}
+	normalized, _ = sjson.SetBytes(normalized, "stream", true)
+	return bytes.Clone(normalized)
+}
+
+func dedupeFunctionCallsByCallID(rawArray string) (string, error) {
+	rawArray = strings.TrimSpace(rawArray)
+	if rawArray == "" {
+		return "[]", nil
+	}
+	var items []json.RawMessage
+	if errUnmarshal := json.Unmarshal([]byte(rawArray), &items); errUnmarshal != nil {
+		return "", errUnmarshal
+	}
+
+	seenCallIDs := make(map[string]struct{}, len(items))
+	filtered := make([]json.RawMessage, 0, len(items))
+	for _, item := range items {
+		if len(item) == 0 {
+			continue
+		}
+		itemType := strings.TrimSpace(gjson.GetBytes(item, "type").String())
+		if isResponsesToolCallType(itemType) {
+			callID := strings.TrimSpace(gjson.GetBytes(item, "call_id").String())
+			if callID != "" {
+				if _, ok := seenCallIDs[callID]; ok {
+					continue
+				}
+				seenCallIDs[callID] = struct{}{}
+			}
+		}
+		filtered = append(filtered, item)
+	}
+
+	out, errMarshal := json.Marshal(filtered)
+	if errMarshal != nil {
+		return "", errMarshal
+	}
+	return string(out), nil
+}
+
 func websocketUpstreamSupportsIncrementalInput(attributes map[string]string, metadata map[string]any) bool {
 	if len(attributes) > 0 {
 		if raw := strings.TrimSpace(attributes["websockets"]); raw != "" {
@@ -492,7 +594,7 @@ func writeResponsesWebsocketSyntheticPrewarm(
 	c *gin.Context,
 	conn *websocket.Conn,
 	requestJSON []byte,
-	wsBodyLog *strings.Builder,
+	wsTimelineLog *strings.Builder,
 	sessionID string,
 ) error {
 	payloads, errPayloads := syntheticResponsesWebsocketPrewarmPayloads(requestJSON)
@@ -501,7 +603,6 @@ func writeResponsesWebsocketSyntheticPrewarm(
 	}
 	for i := 0; i < len(payloads); i++ {
 		markAPIResponseTimestamp(c)
-		appendWebsocketEvent(wsBodyLog, "response", payloads[i])
 		// log.Infof(
 		// 	"responses websocket: downstream_out id=%s type=%d event=%s payload=%s",
 		// 	sessionID,
@@ -509,7 +610,7 @@ func writeResponsesWebsocketSyntheticPrewarm(
 		// 	websocketPayloadEventType(payloads[i]),
 		// 	websocketPayloadPreview(payloads[i]),
 		// )
-		if errWrite := conn.WriteMessage(websocket.TextMessage, payloads[i]); errWrite != nil {
+		if errWrite := writeResponsesWebsocketPayload(conn, wsTimelineLog, payloads[i], time.Now()); errWrite != nil {
 			log.Warnf(
 				"responses websocket: downstream_out write failed id=%s event=%s error=%v",
 				sessionID,
@@ -608,11 +709,15 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesWebsocket(
 	cancel handlers.APIHandlerCancelFunc,
 	data <-chan []byte,
 	errs <-chan *interfaces.ErrorMessage,
-	wsBodyLog *strings.Builder,
+	wsTimelineLog *strings.Builder,
 	sessionID string,
 ) ([]byte, error) {
 	completed := false
 	completedOutput := []byte("[]")
+	downstreamSessionKey := ""
+	if c != nil && c.Request != nil {
+		downstreamSessionKey = websocketDownstreamSessionKey(c.Request)
+	}
 
 	for {
 		select {
@@ -627,8 +732,7 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesWebsocket(
 			if errMsg != nil {
 				h.LoggingAPIResponseError(context.WithValue(context.Background(), "gin", c), errMsg)
 				markAPIResponseTimestamp(c)
-				errorPayload, errWrite := writeResponsesWebsocketError(conn, errMsg)
-				appendWebsocketEvent(wsBodyLog, "response", errorPayload)
+				errorPayload, errWrite := writeResponsesWebsocketError(conn, wsTimelineLog, errMsg)
 				log.Infof(
 					"responses websocket: downstream_out id=%s type=%d event=%s payload=%s",
 					sessionID,
@@ -662,8 +766,7 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesWebsocket(
 					}
 					h.LoggingAPIResponseError(context.WithValue(context.Background(), "gin", c), errMsg)
 					markAPIResponseTimestamp(c)
-					errorPayload, errWrite := writeResponsesWebsocketError(conn, errMsg)
-					appendWebsocketEvent(wsBodyLog, "response", errorPayload)
+					errorPayload, errWrite := writeResponsesWebsocketError(conn, wsTimelineLog, errMsg)
 					log.Infof(
 						"responses websocket: downstream_out id=%s type=%d event=%s payload=%s",
 						sessionID,
@@ -690,13 +793,13 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesWebsocket(
 
 			payloads := websocketJSONPayloadsFromChunk(chunk)
 			for i := range payloads {
+				recordResponsesWebsocketToolCallsFromPayload(downstreamSessionKey, payloads[i])
 				eventType := gjson.GetBytes(payloads[i], "type").String()
 				if eventType == wsEventTypeCompleted {
 					completed = true
 					completedOutput = responseCompletedOutputFromPayload(payloads[i])
 				}
 				markAPIResponseTimestamp(c)
-				appendWebsocketEvent(wsBodyLog, "response", payloads[i])
 				// log.Infof(
 				// 	"responses websocket: downstream_out id=%s type=%d event=%s payload=%s",
 				// 	sessionID,
@@ -704,7 +807,7 @@ func (h *OpenAIResponsesAPIHandler) forwardResponsesWebsocket(
 				// 	websocketPayloadEventType(payloads[i]),
 				// 	websocketPayloadPreview(payloads[i]),
 				// )
-				if errWrite := conn.WriteMessage(websocket.TextMessage, payloads[i]); errWrite != nil {
+				if errWrite := writeResponsesWebsocketPayload(conn, wsTimelineLog, payloads[i], time.Now()); errWrite != nil {
 					log.Warnf(
 						"responses websocket: downstream_out write failed id=%s event=%s error=%v",
 						sessionID,
@@ -760,7 +863,7 @@ func websocketJSONPayloadsFromChunk(chunk []byte) [][]byte {
 	return payloads
 }
 
-func writeResponsesWebsocketError(conn *websocket.Conn, errMsg *interfaces.ErrorMessage) ([]byte, error) {
+func writeResponsesWebsocketError(conn *websocket.Conn, wsTimelineLog *strings.Builder, errMsg *interfaces.ErrorMessage) ([]byte, error) {
 	status := http.StatusInternalServerError
 	errText := http.StatusText(status)
 	if errMsg != nil {
@@ -830,78 +933,25 @@ func writeResponsesWebsocketError(conn *websocket.Conn, errMsg *interfaces.Error
 		}
 	}
 
-	return payload, conn.WriteMessage(websocket.TextMessage, payload)
+	return payload, writeResponsesWebsocketPayload(conn, wsTimelineLog, payload, time.Now())
 }
 
 func appendWebsocketEvent(builder *strings.Builder, eventType string, payload []byte) {
 	if builder == nil {
 		return
 	}
-	if builder.Len() >= wsBodyLogMaxSize {
-		return
-	}
 	trimmedPayload := bytes.TrimSpace(payload)
 	if len(trimmedPayload) == 0 {
 		return
 	}
 	if builder.Len() > 0 {
-		if !appendWebsocketLogString(builder, "\n") {
-			return
-		}
+		builder.WriteString("\n")
 	}
-	if !appendWebsocketLogString(builder, "websocket.") {
-		return
-	}
-	if !appendWebsocketLogString(builder, eventType) {
-		return
-	}
-	if !appendWebsocketLogString(builder, "\n") {
-		return
-	}
-	if !appendWebsocketLogBytes(builder, trimmedPayload, len(wsBodyLogTruncated)) {
-		appendWebsocketLogString(builder, wsBodyLogTruncated)
-		return
-	}
-	appendWebsocketLogString(builder, "\n")
-}
-
-func appendWebsocketLogString(builder *strings.Builder, value string) bool {
-	if builder == nil {
-		return false
-	}
-	remaining := wsBodyLogMaxSize - builder.Len()
-	if remaining <= 0 {
-		return false
-	}
-	if len(value) <= remaining {
-		builder.WriteString(value)
-		return true
-	}
-	builder.WriteString(value[:remaining])
-	return false
-}
-
-func appendWebsocketLogBytes(builder *strings.Builder, value []byte, reserveForSuffix int) bool {
-	if builder == nil {
-		return false
-	}
-	remaining := wsBodyLogMaxSize - builder.Len()
-	if remaining <= 0 {
-		return false
-	}
-	if len(value) <= remaining {
-		builder.Write(value)
-		return true
-	}
-	limit := remaining - reserveForSuffix
-	if limit < 0 {
-		limit = 0
-	}
-	if limit > len(value) {
-		limit = len(value)
-	}
-	builder.Write(value[:limit])
-	return false
+	builder.WriteString("websocket.")
+	builder.WriteString(eventType)
+	builder.WriteString("\n")
+	builder.Write(trimmedPayload)
+	builder.WriteString("\n")
 }
 
 func websocketPayloadEventType(payload []byte) string {
@@ -917,19 +967,16 @@ func websocketPayloadPreview(payload []byte) string {
 	if len(trimmedPayload) == 0 {
 		return "<empty>"
 	}
-	preview := trimmedPayload
-	if len(preview) > wsPayloadLogMaxSize {
-		preview = preview[:wsPayloadLogMaxSize]
-	}
-	previewText := strings.ReplaceAll(string(preview), "\n", "\\n")
+	previewText := strings.ReplaceAll(string(trimmedPayload), "\n", "\\n")
 	previewText = strings.ReplaceAll(previewText, "\r", "\\r")
-	if len(trimmedPayload) > wsPayloadLogMaxSize {
-		return fmt.Sprintf("%s...(truncated,total=%d)", previewText, len(trimmedPayload))
-	}
 	return previewText
 }
 
-func setWebsocketRequestBody(c *gin.Context, body string) {
+func setWebsocketTimelineBody(c *gin.Context, body string) {
+	setWebsocketBody(c, wsTimelineBodyKey, body)
+}
+
+func setWebsocketBody(c *gin.Context, key string, body string) {
 	if c == nil {
 		return
 	}
@@ -937,7 +984,40 @@ func setWebsocketRequestBody(c *gin.Context, body string) {
 	if trimmedBody == "" {
 		return
 	}
-	c.Set(wsRequestBodyKey, []byte(trimmedBody))
+	c.Set(key, []byte(trimmedBody))
+}
+
+func writeResponsesWebsocketPayload(conn *websocket.Conn, wsTimelineLog *strings.Builder, payload []byte, timestamp time.Time) error {
+	appendWebsocketTimelineEvent(wsTimelineLog, "response", payload, timestamp)
+	return conn.WriteMessage(websocket.TextMessage, payload)
+}
+
+func appendWebsocketTimelineDisconnect(builder *strings.Builder, err error, timestamp time.Time) {
+	if err == nil {
+		return
+	}
+	appendWebsocketTimelineEvent(builder, "disconnect", []byte(err.Error()), timestamp)
+}
+
+func appendWebsocketTimelineEvent(builder *strings.Builder, eventType string, payload []byte, timestamp time.Time) {
+	if builder == nil {
+		return
+	}
+	trimmedPayload := bytes.TrimSpace(payload)
+	if len(trimmedPayload) == 0 {
+		return
+	}
+	if builder.Len() > 0 {
+		builder.WriteString("\n")
+	}
+	builder.WriteString("Timestamp: ")
+	builder.WriteString(timestamp.Format(time.RFC3339Nano))
+	builder.WriteString("\n")
+	builder.WriteString("Event: websocket.")
+	builder.WriteString(eventType)
+	builder.WriteString("\n")
+	builder.Write(trimmedPayload)
+	builder.WriteString("\n")
 }
 
 func markAPIResponseTimestamp(c *gin.Context) {

sdk/api/handlers/openai/openai_responses_websocket_test.go

@@ -10,6 +10,7 @@ import (
 	"strings"
 	"sync"
 	"testing"
+	"time"
 
 	"github.com/gin-gonic/gin"
 	"github.com/gorilla/websocket"
@@ -27,6 +28,12 @@ type websocketCaptureExecutor struct {
 	payloads    [][]byte
 }
 
+type websocketCompactionCaptureExecutor struct {
+	mu             sync.Mutex
+	streamPayloads [][]byte
+	compactPayload []byte
+}
+
 type orderedWebsocketSelector struct {
 	mu     sync.Mutex
 	order  []string
@@ -126,6 +133,52 @@ func (e *websocketCaptureExecutor) HttpRequest(context.Context, *coreauth.Auth,
 	return nil, errors.New("not implemented")
 }
 
+func (e *websocketCompactionCaptureExecutor) Identifier() string { return "test-provider" }
+
+func (e *websocketCompactionCaptureExecutor) Execute(_ context.Context, _ *coreauth.Auth, req coreexecutor.Request, opts coreexecutor.Options) (coreexecutor.Response, error) {
+	e.mu.Lock()
+	e.compactPayload = bytes.Clone(req.Payload)
+	e.mu.Unlock()
+	if opts.Alt != "responses/compact" {
+		return coreexecutor.Response{}, fmt.Errorf("unexpected non-compact execute alt: %q", opts.Alt)
+	}
+	return coreexecutor.Response{Payload: []byte(`{"id":"cmp-1","object":"response.compaction"}`)}, nil
+}
+
+func (e *websocketCompactionCaptureExecutor) ExecuteStream(_ context.Context, _ *coreauth.Auth, req coreexecutor.Request, _ coreexecutor.Options) (*coreexecutor.StreamResult, error) {
+	e.mu.Lock()
+	callIndex := len(e.streamPayloads)
+	e.streamPayloads = append(e.streamPayloads, bytes.Clone(req.Payload))
+	e.mu.Unlock()
+
+	var payload []byte
+	switch callIndex {
+	case 0:
+		payload = []byte(`{"type":"response.completed","response":{"id":"resp-1","output":[{"type":"function_call","id":"fc-1","call_id":"call-1","name":"tool"}]}}`)
+	case 1:
+		payload = []byte(`{"type":"response.completed","response":{"id":"resp-2","output":[{"type":"message","id":"assistant-1"}]}}`)
+	default:
+		payload = []byte(`{"type":"response.completed","response":{"id":"resp-3","output":[{"type":"message","id":"assistant-2"}]}}`)
+	}
+
+	chunks := make(chan coreexecutor.StreamChunk, 1)
+	chunks <- coreexecutor.StreamChunk{Payload: payload}
+	close(chunks)
+	return &coreexecutor.StreamResult{Chunks: chunks}, nil
+}
+
+func (e *websocketCompactionCaptureExecutor) Refresh(_ context.Context, auth *coreauth.Auth) (*coreauth.Auth, error) {
+	return auth, nil
+}
+
+func (e *websocketCompactionCaptureExecutor) CountTokens(context.Context, *coreauth.Auth, coreexecutor.Request, coreexecutor.Options) (coreexecutor.Response, error) {
+	return coreexecutor.Response{}, errors.New("not implemented")
+}
+
+func (e *websocketCompactionCaptureExecutor) HttpRequest(context.Context, *coreauth.Auth, *http.Request) (*http.Response, error) {
+	return nil, errors.New("not implemented")
+}
+
 func TestNormalizeResponsesWebsocketRequestCreate(t *testing.T) {
 	raw := []byte(`{"type":"response.create","model":"test-model","stream":false,"input":[{"type":"message","id":"msg-1"}]}`)
 
@@ -339,54 +392,265 @@ func TestAppendWebsocketEvent(t *testing.T) {
 	}
 }
 
-func TestAppendWebsocketEventTruncatesAtLimit(t *testing.T) {
+func TestAppendWebsocketTimelineEvent(t *testing.T) {
 	var builder strings.Builder
-	payload := bytes.Repeat([]byte("x"), wsBodyLogMaxSize)
+	ts := time.Date(2026, time.April, 1, 12, 34, 56, 789000000, time.UTC)
 
-	appendWebsocketEvent(&builder, "request", payload)
+	appendWebsocketTimelineEvent(&builder, "request", []byte("  {\"type\":\"response.create\"}\n"), ts)
 
 	got := builder.String()
-	if len(got) > wsBodyLogMaxSize {
-		t.Fatalf("body log len = %d, want <= %d", len(got), wsBodyLogMaxSize)
+	if !strings.Contains(got, "Timestamp: 2026-04-01T12:34:56.789Z") {
+		t.Fatalf("timeline timestamp not found: %s", got)
 	}
-	if !strings.Contains(got, wsBodyLogTruncated) {
-		t.Fatalf("expected truncation marker in body log")
+	if !strings.Contains(got, "Event: websocket.request") {
+		t.Fatalf("timeline event not found: %s", got)
+	}
+	if !strings.Contains(got, "{\"type\":\"response.create\"}") {
+		t.Fatalf("timeline payload not found: %s", got)
 	}
 }
 
-func TestAppendWebsocketEventNoGrowthAfterLimit(t *testing.T) {
-	var builder strings.Builder
-	appendWebsocketEvent(&builder, "request", bytes.Repeat([]byte("x"), wsBodyLogMaxSize))
-	initial := builder.String()
-
-	appendWebsocketEvent(&builder, "response", []byte(`{"type":"response.completed"}`))
-
-	if builder.String() != initial {
-		t.Fatalf("builder grew after reaching limit")
-	}
-}
-
-func TestSetWebsocketRequestBody(t *testing.T) {
+func TestSetWebsocketTimelineBody(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 	recorder := httptest.NewRecorder()
 	c, _ := gin.CreateTestContext(recorder)
 
-	setWebsocketRequestBody(c, " \n ")
-	if _, exists := c.Get(wsRequestBodyKey); exists {
-		t.Fatalf("request body key should not be set for empty body")
+	setWebsocketTimelineBody(c, " \n ")
+	if _, exists := c.Get(wsTimelineBodyKey); exists {
+		t.Fatalf("timeline body key should not be set for empty body")
 	}
 
-	setWebsocketRequestBody(c, "event body")
-	value, exists := c.Get(wsRequestBodyKey)
+	setWebsocketTimelineBody(c, "timeline body")
+	value, exists := c.Get(wsTimelineBodyKey)
 	if !exists {
-		t.Fatalf("request body key not set")
+		t.Fatalf("timeline body key not set")
 	}
 	bodyBytes, ok := value.([]byte)
 	if !ok {
-		t.Fatalf("request body key type mismatch")
+		t.Fatalf("timeline body key type mismatch")
 	}
-	if string(bodyBytes) != "event body" {
-		t.Fatalf("request body = %q, want %q", string(bodyBytes), "event body")
+	if string(bodyBytes) != "timeline body" {
+		t.Fatalf("timeline body = %q, want %q", string(bodyBytes), "timeline body")
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsInsertsCachedOutput(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	cacheWarm := []byte(`{"previous_response_id":"resp-1","input":[{"type":"function_call_output","call_id":"call-1","output":"ok"}]}`)
+	warmed := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, cacheWarm)
+	if gjson.GetBytes(warmed, "input.0.call_id").String() != "call-1" {
+		t.Fatalf("expected warmup output to remain")
+	}
+
+	raw := []byte(`{"input":[{"type":"function_call","call_id":"call-1","name":"tool"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 3 {
+		t.Fatalf("repaired input len = %d, want 3", len(input))
+	}
+	if input[0].Get("type").String() != "function_call" || input[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("unexpected first item: %s", input[0].Raw)
+	}
+	if input[1].Get("type").String() != "function_call_output" || input[1].Get("call_id").String() != "call-1" {
+		t.Fatalf("missing inserted output: %s", input[1].Raw)
+	}
+	if input[2].Get("type").String() != "message" || input[2].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected trailing item: %s", input[2].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsDropsOrphanFunctionCall(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	raw := []byte(`{"input":[{"type":"function_call","call_id":"call-1","name":"tool"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 1 {
+		t.Fatalf("repaired input len = %d, want 1", len(input))
+	}
+	if input[0].Get("type").String() != "message" || input[0].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected remaining item: %s", input[0].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsInsertsCachedCallForOrphanOutput(t *testing.T) {
+	outputCache := newWebsocketToolOutputCache(time.Minute, 10)
+	callCache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	callCache.record(sessionKey, "call-1", []byte(`{"type":"function_call","call_id":"call-1","name":"tool"}`))
+
+	raw := []byte(`{"input":[{"type":"function_call_output","call_id":"call-1","output":"ok"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCaches(outputCache, callCache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 3 {
+		t.Fatalf("repaired input len = %d, want 3", len(input))
+	}
+	if input[0].Get("type").String() != "function_call" || input[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("missing inserted call: %s", input[0].Raw)
+	}
+	if input[1].Get("type").String() != "function_call_output" || input[1].Get("call_id").String() != "call-1" {
+		t.Fatalf("unexpected output item: %s", input[1].Raw)
+	}
+	if input[2].Get("type").String() != "message" || input[2].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected trailing item: %s", input[2].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsDropsOrphanOutputWhenCallMissing(t *testing.T) {
+	outputCache := newWebsocketToolOutputCache(time.Minute, 10)
+	callCache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	raw := []byte(`{"input":[{"type":"function_call_output","call_id":"call-1","output":"ok"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCaches(outputCache, callCache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 1 {
+		t.Fatalf("repaired input len = %d, want 1", len(input))
+	}
+	if input[0].Get("type").String() != "message" || input[0].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected remaining item: %s", input[0].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsInsertsCachedCustomToolOutput(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	cacheWarm := []byte(`{"previous_response_id":"resp-1","input":[{"type":"custom_tool_call_output","call_id":"call-1","output":"ok"}]}`)
+	warmed := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, cacheWarm)
+	if gjson.GetBytes(warmed, "input.0.call_id").String() != "call-1" {
+		t.Fatalf("expected warmup output to remain")
+	}
+
+	raw := []byte(`{"input":[{"type":"custom_tool_call","call_id":"call-1","name":"apply_patch"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 3 {
+		t.Fatalf("repaired input len = %d, want 3", len(input))
+	}
+	if input[0].Get("type").String() != "custom_tool_call" || input[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("unexpected first item: %s", input[0].Raw)
+	}
+	if input[1].Get("type").String() != "custom_tool_call_output" || input[1].Get("call_id").String() != "call-1" {
+		t.Fatalf("missing inserted output: %s", input[1].Raw)
+	}
+	if input[2].Get("type").String() != "message" || input[2].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected trailing item: %s", input[2].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsDropsOrphanCustomToolCall(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	raw := []byte(`{"input":[{"type":"custom_tool_call","call_id":"call-1","name":"apply_patch"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCache(cache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 1 {
+		t.Fatalf("repaired input len = %d, want 1", len(input))
+	}
+	if input[0].Get("type").String() != "message" || input[0].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected remaining item: %s", input[0].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsInsertsCachedCustomToolCallForOrphanOutput(t *testing.T) {
+	outputCache := newWebsocketToolOutputCache(time.Minute, 10)
+	callCache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	callCache.record(sessionKey, "call-1", []byte(`{"type":"custom_tool_call","call_id":"call-1","name":"apply_patch"}`))
+
+	raw := []byte(`{"input":[{"type":"custom_tool_call_output","call_id":"call-1","output":"ok"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCaches(outputCache, callCache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 3 {
+		t.Fatalf("repaired input len = %d, want 3", len(input))
+	}
+	if input[0].Get("type").String() != "custom_tool_call" || input[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("missing inserted call: %s", input[0].Raw)
+	}
+	if input[1].Get("type").String() != "custom_tool_call_output" || input[1].Get("call_id").String() != "call-1" {
+		t.Fatalf("unexpected output item: %s", input[1].Raw)
+	}
+	if input[2].Get("type").String() != "message" || input[2].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected trailing item: %s", input[2].Raw)
+	}
+}
+
+func TestRepairResponsesWebsocketToolCallsDropsOrphanCustomToolOutputWhenCallMissing(t *testing.T) {
+	outputCache := newWebsocketToolOutputCache(time.Minute, 10)
+	callCache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	raw := []byte(`{"input":[{"type":"custom_tool_call_output","call_id":"call-1","output":"ok"},{"type":"message","id":"msg-1"}]}`)
+	repaired := repairResponsesWebsocketToolCallsWithCaches(outputCache, callCache, sessionKey, raw)
+
+	input := gjson.GetBytes(repaired, "input").Array()
+	if len(input) != 1 {
+		t.Fatalf("repaired input len = %d, want 1", len(input))
+	}
+	if input[0].Get("type").String() != "message" || input[0].Get("id").String() != "msg-1" {
+		t.Fatalf("unexpected remaining item: %s", input[0].Raw)
+	}
+}
+
+func TestRecordResponsesWebsocketToolCallsFromPayloadWithCache(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	payload := []byte(`{"type":"response.completed","response":{"id":"resp-1","output":[{"type":"function_call","id":"fc-1","call_id":"call-1","name":"tool","arguments":"{}"}]}}`)
+	recordResponsesWebsocketToolCallsFromPayloadWithCache(cache, sessionKey, payload)
+
+	cached, ok := cache.get(sessionKey, "call-1")
+	if !ok {
+		t.Fatalf("expected cached tool call")
+	}
+	if gjson.GetBytes(cached, "type").String() != "function_call" || gjson.GetBytes(cached, "call_id").String() != "call-1" {
+		t.Fatalf("unexpected cached tool call: %s", cached)
+	}
+}
+
+func TestRecordResponsesWebsocketCustomToolCallsFromCompletedPayloadWithCache(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	payload := []byte(`{"type":"response.completed","response":{"id":"resp-1","output":[{"type":"custom_tool_call","id":"ctc-1","call_id":"call-1","name":"apply_patch","input":"*** Begin Patch"}]}}`)
+	recordResponsesWebsocketToolCallsFromPayloadWithCache(cache, sessionKey, payload)
+
+	cached, ok := cache.get(sessionKey, "call-1")
+	if !ok {
+		t.Fatalf("expected cached custom tool call")
+	}
+	if gjson.GetBytes(cached, "type").String() != "custom_tool_call" || gjson.GetBytes(cached, "call_id").String() != "call-1" {
+		t.Fatalf("unexpected cached custom tool call: %s", cached)
+	}
+}
+
+func TestRecordResponsesWebsocketCustomToolCallsFromOutputItemDoneWithCache(t *testing.T) {
+	cache := newWebsocketToolOutputCache(time.Minute, 10)
+	sessionKey := "session-1"
+
+	payload := []byte(`{"type":"response.output_item.done","item":{"type":"custom_tool_call","id":"ctc-1","call_id":"call-1","name":"apply_patch","input":"*** Begin Patch"}}`)
+	recordResponsesWebsocketToolCallsFromPayloadWithCache(cache, sessionKey, payload)
+
+	cached, ok := cache.get(sessionKey, "call-1")
+	if !ok {
+		t.Fatalf("expected cached custom tool call")
+	}
+	if gjson.GetBytes(cached, "type").String() != "custom_tool_call" || gjson.GetBytes(cached, "call_id").String() != "call-1" {
+		t.Fatalf("unexpected cached custom tool call: %s", cached)
 	}
 }
 
@@ -416,14 +680,14 @@ func TestForwardResponsesWebsocketPreservesCompletedEvent(t *testing.T) {
 		close(data)
 		close(errCh)
 
-		var bodyLog strings.Builder
+		var timelineLog strings.Builder
 		completedOutput, err := (*OpenAIResponsesAPIHandler)(nil).forwardResponsesWebsocket(
 			ctx,
 			conn,
 			func(...interface{}) {},
 			data,
 			errCh,
-			&bodyLog,
+			&timelineLog,
 			"session-1",
 		)
 		if err != nil {
@@ -434,6 +698,10 @@ func TestForwardResponsesWebsocketPreservesCompletedEvent(t *testing.T) {
 			serverErrCh <- errors.New("completed output not captured")
 			return
 		}
+		if !strings.Contains(timelineLog.String(), "Event: websocket.response") {
+			serverErrCh <- errors.New("websocket timeline did not capture downstream response")
+			return
+		}
 		serverErrCh <- nil
 	}))
 	defer server.Close()
@@ -466,6 +734,116 @@ func TestForwardResponsesWebsocketPreservesCompletedEvent(t *testing.T) {
 	}
 }
 
+func TestForwardResponsesWebsocketLogsAttemptedResponseOnWriteFailure(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	serverErrCh := make(chan error, 1)
+	server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		conn, err := responsesWebsocketUpgrader.Upgrade(w, r, nil)
+		if err != nil {
+			serverErrCh <- err
+			return
+		}
+
+		ctx, _ := gin.CreateTestContext(httptest.NewRecorder())
+		ctx.Request = r
+
+		data := make(chan []byte, 1)
+		errCh := make(chan *interfaces.ErrorMessage)
+		data <- []byte("data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp-1\",\"output\":[{\"type\":\"message\",\"id\":\"out-1\"}]}}\n\n")
+		close(data)
+		close(errCh)
+
+		var timelineLog strings.Builder
+		if errClose := conn.Close(); errClose != nil {
+			serverErrCh <- errClose
+			return
+		}
+
+		_, err = (*OpenAIResponsesAPIHandler)(nil).forwardResponsesWebsocket(
+			ctx,
+			conn,
+			func(...interface{}) {},
+			data,
+			errCh,
+			&timelineLog,
+			"session-1",
+		)
+		if err == nil {
+			serverErrCh <- errors.New("expected websocket write failure")
+			return
+		}
+		if !strings.Contains(timelineLog.String(), "Event: websocket.response") {
+			serverErrCh <- errors.New("websocket timeline did not capture attempted downstream response")
+			return
+		}
+		if !strings.Contains(timelineLog.String(), "\"type\":\"response.completed\"") {
+			serverErrCh <- errors.New("websocket timeline did not retain attempted payload")
+			return
+		}
+		serverErrCh <- nil
+	}))
+	defer server.Close()
+
+	wsURL := "ws" + strings.TrimPrefix(server.URL, "http")
+	conn, _, err := websocket.DefaultDialer.Dial(wsURL, nil)
+	if err != nil {
+		t.Fatalf("dial websocket: %v", err)
+	}
+	defer func() {
+		_ = conn.Close()
+	}()
+
+	if errServer := <-serverErrCh; errServer != nil {
+		t.Fatalf("server error: %v", errServer)
+	}
+}
+
+func TestResponsesWebsocketTimelineRecordsDisconnectEvent(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	manager := coreauth.NewManager(nil, nil, nil)
+	base := handlers.NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, manager)
+	h := NewOpenAIResponsesAPIHandler(base)
+
+	timelineCh := make(chan string, 1)
+	router := gin.New()
+	router.GET("/v1/responses/ws", func(c *gin.Context) {
+		h.ResponsesWebsocket(c)
+		timeline := ""
+		if value, exists := c.Get(wsTimelineBodyKey); exists {
+			if body, ok := value.([]byte); ok {
+				timeline = string(body)
+			}
+		}
+		timelineCh <- timeline
+	})
+
+	server := httptest.NewServer(router)
+	defer server.Close()
+
+	wsURL := "ws" + strings.TrimPrefix(server.URL, "http") + "/v1/responses/ws"
+	conn, _, err := websocket.DefaultDialer.Dial(wsURL, nil)
+	if err != nil {
+		t.Fatalf("dial websocket: %v", err)
+	}
+
+	closePayload := websocket.FormatCloseMessage(websocket.CloseGoingAway, "client closing")
+	if err = conn.WriteControl(websocket.CloseMessage, closePayload, time.Now().Add(time.Second)); err != nil {
+		t.Fatalf("write close control: %v", err)
+	}
+	_ = conn.Close()
+
+	select {
+	case timeline := <-timelineCh:
+		if !strings.Contains(timeline, "Event: websocket.disconnect") {
+			t.Fatalf("websocket timeline missing disconnect event: %s", timeline)
+		}
+	case <-time.After(5 * time.Second):
+		t.Fatal("timed out waiting for websocket timeline")
+	}
+}
+
 func TestWebsocketUpstreamSupportsIncrementalInputForModel(t *testing.T) {
 	manager := coreauth.NewManager(nil, nil, nil)
 	auth := &coreauth.Auth{
@@ -593,6 +971,31 @@ func TestResponsesWebsocketPrewarmHandledLocallyForSSEUpstream(t *testing.T) {
 	}
 }
 
+func TestWebsocketClientAddressUsesGinClientIP(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	recorder := httptest.NewRecorder()
+	c, engine := gin.CreateTestContext(recorder)
+	if err := engine.SetTrustedProxies([]string{"0.0.0.0/0", "::/0"}); err != nil {
+		t.Fatalf("SetTrustedProxies: %v", err)
+	}
+
+	req := httptest.NewRequest(http.MethodGet, "/v1/responses/ws", nil)
+	req.RemoteAddr = "172.18.0.1:34282"
+	req.Header.Set("X-Forwarded-For", "203.0.113.7")
+	c.Request = req
+
+	if got := websocketClientAddress(c); got != strings.TrimSpace(c.ClientIP()) {
+		t.Fatalf("websocketClientAddress = %q, ClientIP = %q", got, c.ClientIP())
+	}
+}
+
+func TestWebsocketClientAddressReturnsEmptyForNilContext(t *testing.T) {
+	if got := websocketClientAddress(nil); got != "" {
+		t.Fatalf("websocketClientAddress(nil) = %q, want empty", got)
+	}
+}
+
 func TestResponsesWebsocketPinsOnlyWebsocketCapableAuth(t *testing.T) {
 	gin.SetMode(gin.TestMode)
 
@@ -662,3 +1065,338 @@ func TestResponsesWebsocketPinsOnlyWebsocketCapableAuth(t *testing.T) {
 		t.Fatalf("selected auth IDs = %v, want [auth-sse auth-ws]", got)
 	}
 }
+
+func TestNormalizeResponsesWebsocketRequestTreatsTranscriptReplacementAsReset(t *testing.T) {
+	lastRequest := []byte(`{"model":"test-model","stream":true,"input":[{"type":"message","id":"msg-1"},{"type":"function_call","id":"fc-1","call_id":"call-1"},{"type":"function_call_output","id":"tool-out-1","call_id":"call-1"},{"type":"message","id":"assistant-1","role":"assistant"}]}`)
+	lastResponseOutput := []byte(`[
+		{"type":"message","id":"assistant-1","role":"assistant"}
+	]`)
+	raw := []byte(`{"type":"response.create","input":[{"type":"function_call","id":"fc-compact","call_id":"call-1","name":"tool"},{"type":"message","id":"msg-2"}]}`)
+
+	normalized, next, errMsg := normalizeResponsesWebsocketRequest(raw, lastRequest, lastResponseOutput)
+	if errMsg != nil {
+		t.Fatalf("unexpected error: %v", errMsg.Error)
+	}
+	if gjson.GetBytes(normalized, "previous_response_id").Exists() {
+		t.Fatalf("previous_response_id must not exist in transcript replacement mode")
+	}
+	items := gjson.GetBytes(normalized, "input").Array()
+	if len(items) != 2 {
+		t.Fatalf("replacement input len = %d, want 2: %s", len(items), normalized)
+	}
+	if items[0].Get("id").String() != "fc-compact" || items[1].Get("id").String() != "msg-2" {
+		t.Fatalf("replacement transcript was not preserved as-is: %s", normalized)
+	}
+	if !bytes.Equal(next, normalized) {
+		t.Fatalf("next request snapshot should match replacement request")
+	}
+}
+
+func TestNormalizeResponsesWebsocketRequestDoesNotTreatDeveloperMessageAsReplacement(t *testing.T) {
+	lastRequest := []byte(`{"model":"test-model","stream":true,"input":[{"type":"message","id":"msg-1"}]}`)
+	lastResponseOutput := []byte(`[
+		{"type":"message","id":"assistant-1","role":"assistant"}
+	]`)
+	raw := []byte(`{"type":"response.create","input":[{"type":"message","id":"dev-1","role":"developer"},{"type":"message","id":"msg-2"}]}`)
+
+	normalized, next, errMsg := normalizeResponsesWebsocketRequest(raw, lastRequest, lastResponseOutput)
+	if errMsg != nil {
+		t.Fatalf("unexpected error: %v", errMsg.Error)
+	}
+	items := gjson.GetBytes(normalized, "input").Array()
+	if len(items) != 4 {
+		t.Fatalf("merged input len = %d, want 4: %s", len(items), normalized)
+	}
+	if items[0].Get("id").String() != "msg-1" ||
+		items[1].Get("id").String() != "assistant-1" ||
+		items[2].Get("id").String() != "dev-1" ||
+		items[3].Get("id").String() != "msg-2" {
+		t.Fatalf("developer follow-up should preserve merge behavior: %s", normalized)
+	}
+	if !bytes.Equal(next, normalized) {
+		t.Fatalf("next request snapshot should match merged request")
+	}
+}
+
+func TestNormalizeResponsesWebsocketRequestDropsDuplicateFunctionCallsByCallID(t *testing.T) {
+	lastRequest := []byte(`{"model":"test-model","stream":true,"input":[{"type":"function_call","id":"fc-1","call_id":"call-1"},{"type":"function_call_output","id":"tool-out-1","call_id":"call-1"}]}`)
+	lastResponseOutput := []byte(`[
+		{"type":"function_call","id":"fc-1","call_id":"call-1","name":"tool"}
+	]`)
+	raw := []byte(`{"type":"response.create","input":[{"type":"message","id":"msg-2"}]}`)
+
+	normalized, _, errMsg := normalizeResponsesWebsocketRequest(raw, lastRequest, lastResponseOutput)
+	if errMsg != nil {
+		t.Fatalf("unexpected error: %v", errMsg.Error)
+	}
+
+	items := gjson.GetBytes(normalized, "input").Array()
+	if len(items) != 3 {
+		t.Fatalf("merged input len = %d, want 3: %s", len(items), normalized)
+	}
+	if items[0].Get("id").String() != "fc-1" ||
+		items[1].Get("id").String() != "tool-out-1" ||
+		items[2].Get("id").String() != "msg-2" {
+		t.Fatalf("unexpected merged input order: %s", normalized)
+	}
+}
+
+func TestNormalizeResponsesWebsocketRequestTreatsCustomToolTranscriptReplacementAsReset(t *testing.T) {
+	lastRequest := []byte(`{"model":"test-model","stream":true,"input":[{"type":"message","id":"msg-1"},{"type":"custom_tool_call","id":"ctc-1","call_id":"call-1","name":"apply_patch"},{"type":"custom_tool_call_output","id":"tool-out-1","call_id":"call-1"},{"type":"message","id":"assistant-1","role":"assistant"}]}`)
+	lastResponseOutput := []byte(`[
+		{"type":"message","id":"assistant-1","role":"assistant"}
+	]`)
+	raw := []byte(`{"type":"response.create","input":[{"type":"custom_tool_call","id":"ctc-compact","call_id":"call-1","name":"apply_patch"},{"type":"custom_tool_call_output","id":"tool-out-compact","call_id":"call-1"},{"type":"message","id":"msg-2"}]}`)
+
+	normalized, next, errMsg := normalizeResponsesWebsocketRequest(raw, lastRequest, lastResponseOutput)
+	if errMsg != nil {
+		t.Fatalf("unexpected error: %v", errMsg.Error)
+	}
+	if gjson.GetBytes(normalized, "previous_response_id").Exists() {
+		t.Fatalf("previous_response_id must not exist in transcript replacement mode")
+	}
+	items := gjson.GetBytes(normalized, "input").Array()
+	if len(items) != 3 {
+		t.Fatalf("replacement input len = %d, want 3: %s", len(items), normalized)
+	}
+	if items[0].Get("id").String() != "ctc-compact" ||
+		items[1].Get("id").String() != "tool-out-compact" ||
+		items[2].Get("id").String() != "msg-2" {
+		t.Fatalf("replacement transcript was not preserved as-is: %s", normalized)
+	}
+	if !bytes.Equal(next, normalized) {
+		t.Fatalf("next request snapshot should match replacement request")
+	}
+}
+
+func TestNormalizeResponsesWebsocketRequestDropsDuplicateCustomToolCallsByCallID(t *testing.T) {
+	lastRequest := []byte(`{"model":"test-model","stream":true,"input":[{"type":"custom_tool_call","id":"ctc-1","call_id":"call-1","name":"apply_patch"},{"type":"custom_tool_call_output","id":"tool-out-1","call_id":"call-1"}]}`)
+	lastResponseOutput := []byte(`[
+		{"type":"custom_tool_call","id":"ctc-1","call_id":"call-1","name":"apply_patch"}
+	]`)
+	raw := []byte(`{"type":"response.create","input":[{"type":"message","id":"msg-2"}]}`)
+
+	normalized, _, errMsg := normalizeResponsesWebsocketRequest(raw, lastRequest, lastResponseOutput)
+	if errMsg != nil {
+		t.Fatalf("unexpected error: %v", errMsg.Error)
+	}
+
+	items := gjson.GetBytes(normalized, "input").Array()
+	if len(items) != 3 {
+		t.Fatalf("merged input len = %d, want 3: %s", len(items), normalized)
+	}
+	if items[0].Get("id").String() != "ctc-1" ||
+		items[1].Get("id").String() != "tool-out-1" ||
+		items[2].Get("id").String() != "msg-2" {
+		t.Fatalf("unexpected merged input order: %s", normalized)
+	}
+}
+
+func TestResponsesWebsocketCompactionResetsTurnStateOnCustomToolTranscriptReplacement(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	executor := &websocketCompactionCaptureExecutor{}
+	manager := coreauth.NewManager(nil, nil, nil)
+	manager.RegisterExecutor(executor)
+	auth := &coreauth.Auth{ID: "auth-sse", Provider: executor.Identifier(), Status: coreauth.StatusActive}
+	if _, err := manager.Register(context.Background(), auth); err != nil {
+		t.Fatalf("Register auth: %v", err)
+	}
+	registry.GetGlobalRegistry().RegisterClient(auth.ID, auth.Provider, []*registry.ModelInfo{{ID: "test-model"}})
+	t.Cleanup(func() {
+		registry.GetGlobalRegistry().UnregisterClient(auth.ID)
+	})
+
+	base := handlers.NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, manager)
+	h := NewOpenAIResponsesAPIHandler(base)
+	router := gin.New()
+	router.GET("/v1/responses/ws", h.ResponsesWebsocket)
+	router.POST("/v1/responses/compact", h.Compact)
+
+	server := httptest.NewServer(router)
+	defer server.Close()
+
+	wsURL := "ws" + strings.TrimPrefix(server.URL, "http") + "/v1/responses/ws"
+	conn, _, err := websocket.DefaultDialer.Dial(wsURL, nil)
+	if err != nil {
+		t.Fatalf("dial websocket: %v", err)
+	}
+	defer func() {
+		if errClose := conn.Close(); errClose != nil {
+			t.Fatalf("close websocket: %v", errClose)
+		}
+	}()
+
+	requests := []string{
+		`{"type":"response.create","model":"test-model","input":[{"type":"message","id":"msg-1"}]}`,
+		`{"type":"response.create","input":[{"type":"custom_tool_call_output","call_id":"call-1","id":"tool-out-1"}]}`,
+	}
+	for i := range requests {
+		if errWrite := conn.WriteMessage(websocket.TextMessage, []byte(requests[i])); errWrite != nil {
+			t.Fatalf("write websocket message %d: %v", i+1, errWrite)
+		}
+		_, payload, errReadMessage := conn.ReadMessage()
+		if errReadMessage != nil {
+			t.Fatalf("read websocket message %d: %v", i+1, errReadMessage)
+		}
+		if got := gjson.GetBytes(payload, "type").String(); got != wsEventTypeCompleted {
+			t.Fatalf("message %d payload type = %s, want %s", i+1, got, wsEventTypeCompleted)
+		}
+	}
+
+	compactResp, errPost := server.Client().Post(
+		server.URL+"/v1/responses/compact",
+		"application/json",
+		strings.NewReader(`{"model":"test-model","input":[{"type":"message","id":"summary-1"}]}`),
+	)
+	if errPost != nil {
+		t.Fatalf("compact request failed: %v", errPost)
+	}
+	if errClose := compactResp.Body.Close(); errClose != nil {
+		t.Fatalf("close compact response body: %v", errClose)
+	}
+	if compactResp.StatusCode != http.StatusOK {
+		t.Fatalf("compact status = %d, want %d", compactResp.StatusCode, http.StatusOK)
+	}
+
+	postCompact := `{"type":"response.create","input":[{"type":"custom_tool_call","id":"ctc-compact","call_id":"call-1","name":"apply_patch"},{"type":"custom_tool_call_output","id":"tool-out-compact","call_id":"call-1"},{"type":"message","id":"msg-2"}]}`
+	if errWrite := conn.WriteMessage(websocket.TextMessage, []byte(postCompact)); errWrite != nil {
+		t.Fatalf("write post-compact websocket message: %v", errWrite)
+	}
+	_, payload, errReadMessage := conn.ReadMessage()
+	if errReadMessage != nil {
+		t.Fatalf("read post-compact websocket message: %v", errReadMessage)
+	}
+	if got := gjson.GetBytes(payload, "type").String(); got != wsEventTypeCompleted {
+		t.Fatalf("post-compact payload type = %s, want %s", got, wsEventTypeCompleted)
+	}
+
+	executor.mu.Lock()
+	defer executor.mu.Unlock()
+
+	if executor.compactPayload == nil {
+		t.Fatalf("compact payload was not captured")
+	}
+	if len(executor.streamPayloads) != 3 {
+		t.Fatalf("stream payload count = %d, want 3", len(executor.streamPayloads))
+	}
+
+	merged := executor.streamPayloads[2]
+	items := gjson.GetBytes(merged, "input").Array()
+	if len(items) != 3 {
+		t.Fatalf("merged input len = %d, want 3: %s", len(items), merged)
+	}
+	if items[0].Get("id").String() != "ctc-compact" ||
+		items[1].Get("id").String() != "tool-out-compact" ||
+		items[2].Get("id").String() != "msg-2" {
+		t.Fatalf("unexpected post-compact input order: %s", merged)
+	}
+	if items[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("post-compact custom tool call id = %s, want call-1", items[0].Get("call_id").String())
+	}
+}
+
+func TestResponsesWebsocketCompactionResetsTurnStateOnTranscriptReplacement(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+
+	executor := &websocketCompactionCaptureExecutor{}
+	manager := coreauth.NewManager(nil, nil, nil)
+	manager.RegisterExecutor(executor)
+	auth := &coreauth.Auth{ID: "auth-sse", Provider: executor.Identifier(), Status: coreauth.StatusActive}
+	if _, err := manager.Register(context.Background(), auth); err != nil {
+		t.Fatalf("Register auth: %v", err)
+	}
+	registry.GetGlobalRegistry().RegisterClient(auth.ID, auth.Provider, []*registry.ModelInfo{{ID: "test-model"}})
+	t.Cleanup(func() {
+		registry.GetGlobalRegistry().UnregisterClient(auth.ID)
+	})
+
+	base := handlers.NewBaseAPIHandlers(&sdkconfig.SDKConfig{}, manager)
+	h := NewOpenAIResponsesAPIHandler(base)
+	router := gin.New()
+	router.GET("/v1/responses/ws", h.ResponsesWebsocket)
+	router.POST("/v1/responses/compact", h.Compact)
+
+	server := httptest.NewServer(router)
+	defer server.Close()
+
+	wsURL := "ws" + strings.TrimPrefix(server.URL, "http") + "/v1/responses/ws"
+	conn, _, err := websocket.DefaultDialer.Dial(wsURL, nil)
+	if err != nil {
+		t.Fatalf("dial websocket: %v", err)
+	}
+	defer func() {
+		if errClose := conn.Close(); errClose != nil {
+			t.Fatalf("close websocket: %v", errClose)
+		}
+	}()
+
+	requests := []string{
+		`{"type":"response.create","model":"test-model","input":[{"type":"message","id":"msg-1"}]}`,
+		`{"type":"response.create","input":[{"type":"function_call_output","call_id":"call-1","id":"tool-out-1"}]}`,
+	}
+	for i := range requests {
+		if errWrite := conn.WriteMessage(websocket.TextMessage, []byte(requests[i])); errWrite != nil {
+			t.Fatalf("write websocket message %d: %v", i+1, errWrite)
+		}
+		_, payload, errReadMessage := conn.ReadMessage()
+		if errReadMessage != nil {
+			t.Fatalf("read websocket message %d: %v", i+1, errReadMessage)
+		}
+		if got := gjson.GetBytes(payload, "type").String(); got != wsEventTypeCompleted {
+			t.Fatalf("message %d payload type = %s, want %s", i+1, got, wsEventTypeCompleted)
+		}
+	}
+
+	compactResp, errPost := server.Client().Post(
+		server.URL+"/v1/responses/compact",
+		"application/json",
+		strings.NewReader(`{"model":"test-model","input":[{"type":"message","id":"summary-1"}]}`),
+	)
+	if errPost != nil {
+		t.Fatalf("compact request failed: %v", errPost)
+	}
+	if errClose := compactResp.Body.Close(); errClose != nil {
+		t.Fatalf("close compact response body: %v", errClose)
+	}
+	if compactResp.StatusCode != http.StatusOK {
+		t.Fatalf("compact status = %d, want %d", compactResp.StatusCode, http.StatusOK)
+	}
+
+	// Simulate a post-compaction client turn that replaces local history with a compacted transcript.
+	// The websocket handler must treat this as a state reset, not append it to stale pre-compaction state.
+	postCompact := `{"type":"response.create","input":[{"type":"function_call","id":"fc-compact","call_id":"call-1","name":"tool"},{"type":"message","id":"msg-2"}]}`
+	if errWrite := conn.WriteMessage(websocket.TextMessage, []byte(postCompact)); errWrite != nil {
+		t.Fatalf("write post-compact websocket message: %v", errWrite)
+	}
+	_, payload, errReadMessage := conn.ReadMessage()
+	if errReadMessage != nil {
+		t.Fatalf("read post-compact websocket message: %v", errReadMessage)
+	}
+	if got := gjson.GetBytes(payload, "type").String(); got != wsEventTypeCompleted {
+		t.Fatalf("post-compact payload type = %s, want %s", got, wsEventTypeCompleted)
+	}
+
+	executor.mu.Lock()
+	defer executor.mu.Unlock()
+
+	if executor.compactPayload == nil {
+		t.Fatalf("compact payload was not captured")
+	}
+	if len(executor.streamPayloads) != 3 {
+		t.Fatalf("stream payload count = %d, want 3", len(executor.streamPayloads))
+	}
+
+	merged := executor.streamPayloads[2]
+	items := gjson.GetBytes(merged, "input").Array()
+	if len(items) != 2 {
+		t.Fatalf("merged input len = %d, want 2: %s", len(items), merged)
+	}
+	if items[0].Get("id").String() != "fc-compact" ||
+		items[1].Get("id").String() != "msg-2" {
+		t.Fatalf("unexpected post-compact input order: %s", merged)
+	}
+	if items[0].Get("call_id").String() != "call-1" {
+		t.Fatalf("post-compact function call id = %s, want call-1", items[0].Get("call_id").String())
+	}
+}

sdk/api/handlers/openai/openai_responses_websocket_toolcall_repair.go

@@ -0,0 +1,420 @@
+package openai
+
+import (
+	"encoding/json"
+	"net/http"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/tidwall/gjson"
+	"github.com/tidwall/sjson"
+)
+
+const (
+	websocketToolOutputCacheMaxPerSession = 256
+	websocketToolOutputCacheTTL           = 30 * time.Minute
+)
+
+var defaultWebsocketToolOutputCache = newWebsocketToolOutputCache(0, websocketToolOutputCacheMaxPerSession)
+var defaultWebsocketToolCallCache = newWebsocketToolOutputCache(0, websocketToolOutputCacheMaxPerSession)
+var defaultWebsocketToolSessionRefs = newWebsocketToolSessionRefCounter()
+
+type websocketToolOutputCache struct {
+	mu            sync.Mutex
+	ttl           time.Duration
+	maxPerSession int
+	sessions      map[string]*websocketToolOutputSession
+}
+
+type websocketToolOutputSession struct {
+	lastSeen time.Time
+	outputs  map[string]json.RawMessage
+	order    []string
+}
+
+func newWebsocketToolOutputCache(ttl time.Duration, maxPerSession int) *websocketToolOutputCache {
+	if ttl < 0 {
+		ttl = websocketToolOutputCacheTTL
+	}
+	if maxPerSession <= 0 {
+		maxPerSession = websocketToolOutputCacheMaxPerSession
+	}
+	return &websocketToolOutputCache{
+		ttl:           ttl,
+		maxPerSession: maxPerSession,
+		sessions:      make(map[string]*websocketToolOutputSession),
+	}
+}
+
+func (c *websocketToolOutputCache) record(sessionKey string, callID string, item json.RawMessage) {
+	sessionKey = strings.TrimSpace(sessionKey)
+	callID = strings.TrimSpace(callID)
+	if sessionKey == "" || callID == "" || c == nil {
+		return
+	}
+
+	now := time.Now()
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	c.cleanupLocked(now)
+
+	session, ok := c.sessions[sessionKey]
+	if !ok || session == nil {
+		session = &websocketToolOutputSession{
+			lastSeen: now,
+			outputs:  make(map[string]json.RawMessage),
+		}
+		c.sessions[sessionKey] = session
+	}
+	session.lastSeen = now
+
+	if _, exists := session.outputs[callID]; !exists {
+		session.order = append(session.order, callID)
+	}
+	session.outputs[callID] = append(json.RawMessage(nil), item...)
+
+	for len(session.order) > c.maxPerSession {
+		evict := session.order[0]
+		session.order = session.order[1:]
+		delete(session.outputs, evict)
+	}
+}
+
+func (c *websocketToolOutputCache) get(sessionKey string, callID string) (json.RawMessage, bool) {
+	sessionKey = strings.TrimSpace(sessionKey)
+	callID = strings.TrimSpace(callID)
+	if sessionKey == "" || callID == "" || c == nil {
+		return nil, false
+	}
+
+	now := time.Now()
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	c.cleanupLocked(now)
+
+	session, ok := c.sessions[sessionKey]
+	if !ok || session == nil {
+		return nil, false
+	}
+	session.lastSeen = now
+	item, ok := session.outputs[callID]
+	if !ok || len(item) == 0 {
+		return nil, false
+	}
+	return append(json.RawMessage(nil), item...), true
+}
+
+func (c *websocketToolOutputCache) cleanupLocked(now time.Time) {
+	if c == nil || c.ttl <= 0 {
+		return
+	}
+
+	for key, session := range c.sessions {
+		if session == nil {
+			delete(c.sessions, key)
+			continue
+		}
+		if now.Sub(session.lastSeen) > c.ttl {
+			delete(c.sessions, key)
+		}
+	}
+}
+
+func (c *websocketToolOutputCache) deleteSession(sessionKey string) {
+	sessionKey = strings.TrimSpace(sessionKey)
+	if sessionKey == "" || c == nil {
+		return
+	}
+
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	delete(c.sessions, sessionKey)
+}
+
+func websocketDownstreamSessionKey(req *http.Request) string {
+	if req == nil {
+		return ""
+	}
+	if requestID := strings.TrimSpace(req.Header.Get("X-Client-Request-Id")); requestID != "" {
+		return requestID
+	}
+	if raw := strings.TrimSpace(req.Header.Get("X-Codex-Turn-Metadata")); raw != "" {
+		if sessionID := strings.TrimSpace(gjson.Get(raw, "session_id").String()); sessionID != "" {
+			return sessionID
+		}
+	}
+	if sessionID := strings.TrimSpace(req.Header.Get("Session_id")); sessionID != "" {
+		return sessionID
+	}
+	return ""
+}
+
+type websocketToolSessionRefCounter struct {
+	mu     sync.Mutex
+	counts map[string]int
+}
+
+func newWebsocketToolSessionRefCounter() *websocketToolSessionRefCounter {
+	return &websocketToolSessionRefCounter{counts: make(map[string]int)}
+}
+
+func (c *websocketToolSessionRefCounter) acquire(sessionKey string) {
+	sessionKey = strings.TrimSpace(sessionKey)
+	if sessionKey == "" || c == nil {
+		return
+	}
+
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	c.counts[sessionKey]++
+}
+
+func (c *websocketToolSessionRefCounter) release(sessionKey string) bool {
+	sessionKey = strings.TrimSpace(sessionKey)
+	if sessionKey == "" || c == nil {
+		return false
+	}
+
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	count := c.counts[sessionKey]
+	if count <= 1 {
+		delete(c.counts, sessionKey)
+		return true
+	}
+	c.counts[sessionKey] = count - 1
+	return false
+}
+
+func retainResponsesWebsocketToolCaches(sessionKey string) {
+	if defaultWebsocketToolSessionRefs == nil {
+		return
+	}
+	defaultWebsocketToolSessionRefs.acquire(sessionKey)
+}
+
+func releaseResponsesWebsocketToolCaches(sessionKey string) {
+	if defaultWebsocketToolSessionRefs == nil {
+		return
+	}
+	if !defaultWebsocketToolSessionRefs.release(sessionKey) {
+		return
+	}
+
+	if defaultWebsocketToolOutputCache != nil {
+		defaultWebsocketToolOutputCache.deleteSession(sessionKey)
+	}
+	if defaultWebsocketToolCallCache != nil {
+		defaultWebsocketToolCallCache.deleteSession(sessionKey)
+	}
+}
+
+func repairResponsesWebsocketToolCalls(sessionKey string, payload []byte) []byte {
+	return repairResponsesWebsocketToolCallsWithCaches(defaultWebsocketToolOutputCache, defaultWebsocketToolCallCache, sessionKey, payload)
+}
+
+func repairResponsesWebsocketToolCallsWithCache(cache *websocketToolOutputCache, sessionKey string, payload []byte) []byte {
+	return repairResponsesWebsocketToolCallsWithCaches(cache, nil, sessionKey, payload)
+}
+
+func repairResponsesWebsocketToolCallsWithCaches(outputCache, callCache *websocketToolOutputCache, sessionKey string, payload []byte) []byte {
+	sessionKey = strings.TrimSpace(sessionKey)
+	if sessionKey == "" || outputCache == nil || len(payload) == 0 {
+		return payload
+	}
+
+	input := gjson.GetBytes(payload, "input")
+	if !input.Exists() || !input.IsArray() {
+		return payload
+	}
+
+	allowOrphanOutputs := strings.TrimSpace(gjson.GetBytes(payload, "previous_response_id").String()) != ""
+	updatedRaw, errRepair := repairResponsesToolCallsArray(outputCache, callCache, sessionKey, input.Raw, allowOrphanOutputs)
+	if errRepair != nil || updatedRaw == "" || updatedRaw == input.Raw {
+		return payload
+	}
+
+	updated, errSet := sjson.SetRawBytes(payload, "input", []byte(updatedRaw))
+	if errSet != nil {
+		return payload
+	}
+	return updated
+}
+
+func repairResponsesToolCallsArray(outputCache, callCache *websocketToolOutputCache, sessionKey string, rawArray string, allowOrphanOutputs bool) (string, error) {
+	rawArray = strings.TrimSpace(rawArray)
+	if rawArray == "" {
+		return "[]", nil
+	}
+
+	var items []json.RawMessage
+	if errUnmarshal := json.Unmarshal([]byte(rawArray), &items); errUnmarshal != nil {
+		return "", errUnmarshal
+	}
+
+	// First pass: record tool outputs and remember which call_ids have outputs in this payload.
+	outputPresent := make(map[string]struct{}, len(items))
+	callPresent := make(map[string]struct{}, len(items))
+	for _, item := range items {
+		if len(item) == 0 {
+			continue
+		}
+		itemType := strings.TrimSpace(gjson.GetBytes(item, "type").String())
+		switch {
+		case isResponsesToolCallOutputType(itemType):
+			callID := strings.TrimSpace(gjson.GetBytes(item, "call_id").String())
+			if callID == "" {
+				continue
+			}
+			outputPresent[callID] = struct{}{}
+			outputCache.record(sessionKey, callID, item)
+		case isResponsesToolCallType(itemType):
+			callID := strings.TrimSpace(gjson.GetBytes(item, "call_id").String())
+			if callID == "" {
+				continue
+			}
+			callPresent[callID] = struct{}{}
+			if callCache != nil {
+				callCache.record(sessionKey, callID, item)
+			}
+		}
+	}
+
+	filtered := make([]json.RawMessage, 0, len(items))
+	insertedCalls := make(map[string]struct{}, len(items))
+	for _, item := range items {
+		if len(item) == 0 {
+			continue
+		}
+		itemType := strings.TrimSpace(gjson.GetBytes(item, "type").String())
+		if isResponsesToolCallOutputType(itemType) {
+			callID := strings.TrimSpace(gjson.GetBytes(item, "call_id").String())
+			if callID == "" {
+				// Upstream rejects tool outputs without a call_id; drop it.
+				continue
+			}
+
+			if allowOrphanOutputs {
+				filtered = append(filtered, item)
+				continue
+			}
+
+			if _, ok := callPresent[callID]; ok {
+				filtered = append(filtered, item)
+				continue
+			}
+
+			if callCache != nil {
+				if cached, ok := callCache.get(sessionKey, callID); ok {
+					if _, already := insertedCalls[callID]; !already {
+						filtered = append(filtered, cached)
+						insertedCalls[callID] = struct{}{}
+						callPresent[callID] = struct{}{}
+					}
+					filtered = append(filtered, item)
+					continue
+				}
+			}
+
+			// Drop orphaned function_call_output items; upstream rejects transcripts with missing calls.
+			continue
+		}
+		if !isResponsesToolCallType(itemType) {
+			filtered = append(filtered, item)
+			continue
+		}
+
+		callID := strings.TrimSpace(gjson.GetBytes(item, "call_id").String())
+		if callID == "" {
+			// Upstream rejects tool calls without a call_id; drop it.
+			continue
+		}
+
+		if _, ok := outputPresent[callID]; ok {
+			filtered = append(filtered, item)
+			continue
+		}
+
+		if cached, ok := outputCache.get(sessionKey, callID); ok {
+			filtered = append(filtered, item)
+			filtered = append(filtered, cached)
+			outputPresent[callID] = struct{}{}
+			continue
+		}
+
+		// Drop orphaned function_call items; upstream rejects transcripts with missing outputs.
+	}
+
+	out, errMarshal := json.Marshal(filtered)
+	if errMarshal != nil {
+		return "", errMarshal
+	}
+	return string(out), nil
+}
+
+func recordResponsesWebsocketToolCallsFromPayload(sessionKey string, payload []byte) {
+	recordResponsesWebsocketToolCallsFromPayloadWithCache(defaultWebsocketToolCallCache, sessionKey, payload)
+}
+
+func recordResponsesWebsocketToolCallsFromPayloadWithCache(cache *websocketToolOutputCache, sessionKey string, payload []byte) {
+	sessionKey = strings.TrimSpace(sessionKey)
+	if sessionKey == "" || cache == nil || len(payload) == 0 {
+		return
+	}
+
+	eventType := strings.TrimSpace(gjson.GetBytes(payload, "type").String())
+	switch eventType {
+	case "response.completed":
+		output := gjson.GetBytes(payload, "response.output")
+		if !output.Exists() || !output.IsArray() {
+			return
+		}
+		for _, item := range output.Array() {
+			if !isResponsesToolCallType(item.Get("type").String()) {
+				continue
+			}
+			callID := strings.TrimSpace(item.Get("call_id").String())
+			if callID == "" {
+				continue
+			}
+			cache.record(sessionKey, callID, json.RawMessage(item.Raw))
+		}
+	case "response.output_item.added", "response.output_item.done":
+		item := gjson.GetBytes(payload, "item")
+		if !item.Exists() || !item.IsObject() {
+			return
+		}
+		if !isResponsesToolCallType(item.Get("type").String()) {
+			return
+		}
+		callID := strings.TrimSpace(item.Get("call_id").String())
+		if callID == "" {
+			return
+		}
+		cache.record(sessionKey, callID, json.RawMessage(item.Raw))
+	}
+}
+
+func isResponsesToolCallType(itemType string) bool {
+	switch strings.TrimSpace(itemType) {
+	case "function_call", "custom_tool_call":
+		return true
+	default:
+		return false
+	}
+}
+
+func isResponsesToolCallOutputType(itemType string) bool {
+	switch strings.TrimSpace(itemType) {
+	case "function_call_output", "custom_tool_call_output":
+		return true
+	default:
+		return false
+	}
+}

sdk/auth/filestore.go

@@ -263,6 +263,7 @@ func (s *FileTokenStore) readAuthFile(path, baseDir string) (*cliproxyauth.Auth,
 	if email, ok := metadata["email"].(string); ok && email != "" {
 		auth.Attributes["email"] = email
 	}
+	cliproxyauth.ApplyCustomHeadersFromMetadata(auth)
 	return auth, nil
 }
 

sdk/cliproxy/auth/conductor.go

@@ -8,6 +8,7 @@ import (
 	"io"
 	"net/http"
 	"path/filepath"
+	"sort"
 	"strconv"
 	"strings"
 	"sync"
@@ -437,6 +438,31 @@ func (m *Manager) executionModelCandidates(auth *Auth, routeModel string) []stri
 	return []string{resolved}
 }
 
+func (m *Manager) selectionModelForAuth(auth *Auth, routeModel string) string {
+	requestedModel := rewriteModelForAuth(routeModel, auth)
+	if strings.TrimSpace(requestedModel) == "" {
+		requestedModel = strings.TrimSpace(routeModel)
+	}
+	resolvedModel := m.applyOAuthModelAlias(auth, requestedModel)
+	if strings.TrimSpace(resolvedModel) == "" {
+		resolvedModel = requestedModel
+	}
+	return resolvedModel
+}
+
+func (m *Manager) selectionModelKeyForAuth(auth *Auth, routeModel string) string {
+	return canonicalModelKey(m.selectionModelForAuth(auth, routeModel))
+}
+
+func (m *Manager) stateModelForExecution(auth *Auth, routeModel, upstreamModel string, pooled bool) string {
+	stateModel := executionResultModel(routeModel, upstreamModel, pooled)
+	selectionModel := m.selectionModelForAuth(auth, routeModel)
+	if canonicalModelKey(selectionModel) == canonicalModelKey(upstreamModel) && strings.TrimSpace(selectionModel) != "" {
+		return strings.TrimSpace(upstreamModel)
+	}
+	return stateModel
+}
+
 func executionResultModel(routeModel, upstreamModel string, pooled bool) string {
 	if pooled {
 		if resolved := strings.TrimSpace(upstreamModel); resolved != "" {
@@ -449,14 +475,14 @@ func executionResultModel(routeModel, upstreamModel string, pooled bool) string
 	return strings.TrimSpace(upstreamModel)
 }
 
-func filterExecutionModels(auth *Auth, routeModel string, candidates []string, pooled bool) []string {
+func (m *Manager) filterExecutionModels(auth *Auth, routeModel string, candidates []string, pooled bool) []string {
 	if len(candidates) == 0 {
 		return nil
 	}
 	now := time.Now()
 	out := make([]string, 0, len(candidates))
 	for _, upstreamModel := range candidates {
-		stateModel := executionResultModel(routeModel, upstreamModel, pooled)
+		stateModel := m.stateModelForExecution(auth, routeModel, upstreamModel, pooled)
 		blocked, _, _ := isAuthBlockedForModel(auth, stateModel, now)
 		if blocked {
 			continue
@@ -469,7 +495,7 @@ func filterExecutionModels(auth *Auth, routeModel string, candidates []string, p
 func (m *Manager) preparedExecutionModels(auth *Auth, routeModel string) ([]string, bool) {
 	candidates := m.executionModelCandidates(auth, routeModel)
 	pooled := len(candidates) > 1
-	return filterExecutionModels(auth, routeModel, candidates, pooled), pooled
+	return m.filterExecutionModels(auth, routeModel, candidates, pooled), pooled
 }
 
 func (m *Manager) prepareExecutionModels(auth *Auth, routeModel string) []string {
@@ -477,6 +503,83 @@ func (m *Manager) prepareExecutionModels(auth *Auth, routeModel string) []string
 	return models
 }
 
+func (m *Manager) availableAuthsForRouteModel(auths []*Auth, provider, routeModel string, now time.Time) ([]*Auth, error) {
+	if len(auths) == 0 {
+		return nil, &Error{Code: "auth_not_found", Message: "no auth candidates"}
+	}
+
+	availableByPriority := make(map[int][]*Auth)
+	cooldownCount := 0
+	var earliest time.Time
+	for _, candidate := range auths {
+		checkModel := m.selectionModelForAuth(candidate, routeModel)
+		blocked, reason, next := isAuthBlockedForModel(candidate, checkModel, now)
+		if !blocked {
+			priority := authPriority(candidate)
+			availableByPriority[priority] = append(availableByPriority[priority], candidate)
+			continue
+		}
+		if reason == blockReasonCooldown {
+			cooldownCount++
+			if !next.IsZero() && (earliest.IsZero() || next.Before(earliest)) {
+				earliest = next
+			}
+		}
+	}
+
+	if len(availableByPriority) == 0 {
+		if cooldownCount == len(auths) && !earliest.IsZero() {
+			providerForError := provider
+			if providerForError == "mixed" {
+				providerForError = ""
+			}
+			resetIn := earliest.Sub(now)
+			if resetIn < 0 {
+				resetIn = 0
+			}
+			return nil, newModelCooldownError(routeModel, providerForError, resetIn)
+		}
+		return nil, &Error{Code: "auth_unavailable", Message: "no auth available"}
+	}
+
+	bestPriority := 0
+	found := false
+	for priority := range availableByPriority {
+		if !found || priority > bestPriority {
+			bestPriority = priority
+			found = true
+		}
+	}
+
+	available := availableByPriority[bestPriority]
+	if len(available) > 1 {
+		sort.Slice(available, func(i, j int) bool { return available[i].ID < available[j].ID })
+	}
+	return available, nil
+}
+
+func selectionArgForSelector(selector Selector, routeModel string) string {
+	if isBuiltInSelector(selector) {
+		return ""
+	}
+	return routeModel
+}
+
+func (m *Manager) authSupportsRouteModel(registryRef *registry.ModelRegistry, auth *Auth, routeModel string) bool {
+	if registryRef == nil || auth == nil {
+		return true
+	}
+	routeKey := canonicalModelKey(routeModel)
+	if routeKey == "" {
+		return true
+	}
+	if registryRef.ClientSupportsModel(auth.ID, routeKey) {
+		return true
+	}
+	selectionKey := m.selectionModelKeyForAuth(auth, routeModel)
+	return selectionKey != "" && selectionKey != routeKey && registryRef.ClientSupportsModel(auth.ID, selectionKey)
+}
+
 func discardStreamChunks(ch <-chan cliproxyexecutor.StreamChunk) {
 	if ch == nil {
 		return
@@ -627,7 +730,7 @@ func (m *Manager) executeStreamWithModelPool(ctx context.Context, executor Provi
 	}
 	var lastErr error
 	for idx, execModel := range execModels {
-		resultModel := executionResultModel(routeModel, execModel, pooled)
+		resultModel := m.stateModelForExecution(auth, routeModel, execModel, pooled)
 		execReq := req
 		execReq.Model = execModel
 		streamResult, errStream := executor.ExecuteStream(ctx, auth, execReq, opts)
@@ -923,8 +1026,10 @@ func (m *Manager) Update(ctx context.Context, auth *Auth) (*Auth, error) {
 			auth.Index = existing.Index
 			auth.indexAssigned = existing.indexAssigned
 		}
-		if len(auth.ModelStates) == 0 && len(existing.ModelStates) > 0 {
-			auth.ModelStates = existing.ModelStates
+		if !existing.Disabled && existing.Status != StatusDisabled && !auth.Disabled && auth.Status != StatusDisabled {
+			if len(auth.ModelStates) == 0 && len(existing.ModelStates) > 0 {
+				auth.ModelStates = existing.ModelStates
+			}
 		}
 	}
 	auth.EnsureIndex()
@@ -1105,7 +1210,7 @@ func (m *Manager) executeMixedOnce(ctx context.Context, providers []string, req
 		attempted[auth.ID] = struct{}{}
 		var authErr error
 		for _, upstreamModel := range models {
-			resultModel := executionResultModel(routeModel, upstreamModel, pooled)
+			resultModel := m.stateModelForExecution(auth, routeModel, upstreamModel, pooled)
 			execReq := req
 			execReq.Model = upstreamModel
 			resp, errExec := executor.Execute(execCtx, auth, execReq, opts)
@@ -1183,7 +1288,7 @@ func (m *Manager) executeCountMixedOnce(ctx context.Context, providers []string,
 		attempted[auth.ID] = struct{}{}
 		var authErr error
 		for _, upstreamModel := range models {
-			resultModel := executionResultModel(routeModel, upstreamModel, pooled)
+			resultModel := m.stateModelForExecution(auth, routeModel, upstreamModel, pooled)
 			execReq := req
 			execReq.Model = upstreamModel
 			resp, errExec := executor.CountTokens(execCtx, auth, execReq, opts)
@@ -1732,77 +1837,79 @@ func (m *Manager) MarkResult(ctx context.Context, result Result) {
 			}
 		} else {
 			if result.Model != "" {
-				state := ensureModelState(auth, result.Model)
-				state.Unavailable = true
-				state.Status = StatusError
-				state.UpdatedAt = now
-				if result.Error != nil {
-					state.LastError = cloneError(result.Error)
-					state.StatusMessage = result.Error.Message
-					auth.LastError = cloneError(result.Error)
-					auth.StatusMessage = result.Error.Message
-				}
+				if !isRequestScopedNotFoundResultError(result.Error) {
+					state := ensureModelState(auth, result.Model)
+					state.Unavailable = true
+					state.Status = StatusError
+					state.UpdatedAt = now
+					if result.Error != nil {
+						state.LastError = cloneError(result.Error)
+						state.StatusMessage = result.Error.Message
+						auth.LastError = cloneError(result.Error)
+						auth.StatusMessage = result.Error.Message
+					}
 
-				statusCode := statusCodeFromResult(result.Error)
-				if isModelSupportResultError(result.Error) {
-					next := now.Add(12 * time.Hour)
-					state.NextRetryAfter = next
-					suspendReason = "model_not_supported"
-					shouldSuspendModel = true
-				} else {
-					switch statusCode {
-					case 401:
-						next := now.Add(30 * time.Minute)
-						state.NextRetryAfter = next
-						suspendReason = "unauthorized"
-						shouldSuspendModel = true
-					case 402, 403:
-						next := now.Add(30 * time.Minute)
-						state.NextRetryAfter = next
-						suspendReason = "payment_required"
-						shouldSuspendModel = true
-					case 404:
+					statusCode := statusCodeFromResult(result.Error)
+					if isModelSupportResultError(result.Error) {
 						next := now.Add(12 * time.Hour)
 						state.NextRetryAfter = next
-						suspendReason = "not_found"
+						suspendReason = "model_not_supported"
 						shouldSuspendModel = true
-					case 429:
-						var next time.Time
-						backoffLevel := state.Quota.BackoffLevel
-						if result.RetryAfter != nil {
-							next = now.Add(*result.RetryAfter)
-						} else {
-							cooldown, nextLevel := nextQuotaCooldown(backoffLevel, quotaCooldownDisabledForAuth(auth))
-							if cooldown > 0 {
-								next = now.Add(cooldown)
-							}
-							backoffLevel = nextLevel
-						}
-						state.NextRetryAfter = next
-						state.Quota = QuotaState{
-							Exceeded:      true,
-							Reason:        "quota",
-							NextRecoverAt: next,
-							BackoffLevel:  backoffLevel,
-						}
-						suspendReason = "quota"
-						shouldSuspendModel = true
-						setModelQuota = true
-					case 408, 500, 502, 503, 504:
-						if quotaCooldownDisabledForAuth(auth) {
-							state.NextRetryAfter = time.Time{}
-						} else {
-							next := now.Add(1 * time.Minute)
+					} else {
+						switch statusCode {
+						case 401:
+							next := now.Add(30 * time.Minute)
 							state.NextRetryAfter = next
+							suspendReason = "unauthorized"
+							shouldSuspendModel = true
+						case 402, 403:
+							next := now.Add(30 * time.Minute)
+							state.NextRetryAfter = next
+							suspendReason = "payment_required"
+							shouldSuspendModel = true
+						case 404:
+							next := now.Add(12 * time.Hour)
+							state.NextRetryAfter = next
+							suspendReason = "not_found"
+							shouldSuspendModel = true
+						case 429:
+							var next time.Time
+							backoffLevel := state.Quota.BackoffLevel
+							if result.RetryAfter != nil {
+								next = now.Add(*result.RetryAfter)
+							} else {
+								cooldown, nextLevel := nextQuotaCooldown(backoffLevel, quotaCooldownDisabledForAuth(auth))
+								if cooldown > 0 {
+									next = now.Add(cooldown)
+								}
+								backoffLevel = nextLevel
+							}
+							state.NextRetryAfter = next
+							state.Quota = QuotaState{
+								Exceeded:      true,
+								Reason:        "quota",
+								NextRecoverAt: next,
+								BackoffLevel:  backoffLevel,
+							}
+							suspendReason = "quota"
+							shouldSuspendModel = true
+							setModelQuota = true
+						case 408, 500, 502, 503, 504:
+							if quotaCooldownDisabledForAuth(auth) {
+								state.NextRetryAfter = time.Time{}
+							} else {
+								next := now.Add(1 * time.Minute)
+								state.NextRetryAfter = next
+							}
+						default:
+							state.NextRetryAfter = time.Time{}
 						}
-					default:
-						state.NextRetryAfter = time.Time{}
 					}
-				}
 
-				auth.Status = StatusError
-				auth.UpdatedAt = now
-				updateAggregatedAvailability(auth, now)
+					auth.Status = StatusError
+					auth.UpdatedAt = now
+					updateAggregatedAvailability(auth, now)
+				}
 			} else {
 				applyAuthFailureState(auth, result.Error, result.RetryAfter, now)
 			}
@@ -2054,11 +2161,29 @@ func isModelSupportResultError(err *Error) bool {
 	return isModelSupportErrorMessage(err.Message)
 }
 
+func isRequestScopedNotFoundMessage(message string) bool {
+	if message == "" {
+		return false
+	}
+	lower := strings.ToLower(message)
+	return strings.Contains(lower, "item with id") &&
+		strings.Contains(lower, "not found") &&
+		strings.Contains(lower, "items are not persisted when `store` is set to false")
+}
+
+func isRequestScopedNotFoundResultError(err *Error) bool {
+	if err == nil || statusCodeFromResult(err) != http.StatusNotFound {
+		return false
+	}
+	return isRequestScopedNotFoundMessage(err.Message)
+}
+
 // isRequestInvalidError returns true if the error represents a client request
 // error that should not be retried. Specifically, it treats 400 responses with
-// "invalid_request_error" and all 422 responses as request-shape failures,
-// where switching auths or pooled upstream models will not help. Model-support
-// errors are excluded so routing can fall through to another auth or upstream.
+// "invalid_request_error", request-scoped 404 item misses caused by `store=false`,
+// and all 422 responses as request-shape failures, where switching auths or
+// pooled upstream models will not help. Model-support errors are excluded so
+// routing can fall through to another auth or upstream.
 func isRequestInvalidError(err error) bool {
 	if err == nil {
 		return false
@@ -2070,6 +2195,8 @@ func isRequestInvalidError(err error) bool {
 	switch status {
 	case http.StatusBadRequest:
 		return strings.Contains(err.Error(), "invalid_request_error")
+	case http.StatusNotFound:
+		return isRequestScopedNotFoundMessage(err.Error())
 	case http.StatusUnprocessableEntity:
 		return true
 	default:
@@ -2081,6 +2208,9 @@ func applyAuthFailureState(auth *Auth, resultErr *Error, retryAfter *time.Durati
 	if auth == nil {
 		return
 	}
+	if isRequestScopedNotFoundResultError(resultErr) {
+		return
+	}
 	auth.Unavailable = true
 	auth.Status = StatusError
 	auth.UpdatedAt = now
@@ -2244,6 +2374,13 @@ func shouldRetrySchedulerPick(err error) bool {
 	return authErr.Code == "auth_not_found" || authErr.Code == "auth_unavailable"
 }
 
+func (m *Manager) routeAwareSelectionRequired(auth *Auth, routeModel string) bool {
+	if auth == nil || strings.TrimSpace(routeModel) == "" {
+		return false
+	}
+	return m.selectionModelKeyForAuth(auth, routeModel) != canonicalModelKey(routeModel)
+}
+
 func (m *Manager) pickNextLegacy(ctx context.Context, provider, model string, opts cliproxyexecutor.Options, tried map[string]struct{}) (*Auth, ProviderExecutor, error) {
 	pinnedAuthID := pinnedAuthIDFromMetadata(opts.Metadata)
 
@@ -2273,7 +2410,7 @@ func (m *Manager) pickNextLegacy(ctx context.Context, provider, model string, op
 		if _, used := tried[candidate.ID]; used {
 			continue
 		}
-		if modelKey != "" && registryRef != nil && !registryRef.ClientSupportsModel(candidate.ID, modelKey) {
+		if modelKey != "" && !m.authSupportsRouteModel(registryRef, candidate, model) {
 			continue
 		}
 		candidates = append(candidates, candidate)
@@ -2282,7 +2419,12 @@ func (m *Manager) pickNextLegacy(ctx context.Context, provider, model string, op
 		m.mu.RUnlock()
 		return nil, nil, &Error{Code: "auth_not_found", Message: "no auth available"}
 	}
-	selected, errPick := m.selector.Pick(ctx, provider, model, opts, candidates)
+	available, errAvailable := m.availableAuthsForRouteModel(candidates, provider, model, time.Now())
+	if errAvailable != nil {
+		m.mu.RUnlock()
+		return nil, nil, errAvailable
+	}
+	selected, errPick := m.selector.Pick(ctx, provider, selectionArgForSelector(m.selector, model), opts, available)
 	if errPick != nil {
 		m.mu.RUnlock()
 		return nil, nil, errPick
@@ -2308,6 +2450,22 @@ func (m *Manager) pickNext(ctx context.Context, provider, model string, opts cli
 	if !m.useSchedulerFastPath() {
 		return m.pickNextLegacy(ctx, provider, model, opts, tried)
 	}
+	if strings.TrimSpace(model) != "" {
+		m.mu.RLock()
+		for _, candidate := range m.auths {
+			if candidate == nil || candidate.Provider != provider || candidate.Disabled {
+				continue
+			}
+			if _, used := tried[candidate.ID]; used {
+				continue
+			}
+			if m.routeAwareSelectionRequired(candidate, model) {
+				m.mu.RUnlock()
+				return m.pickNextLegacy(ctx, provider, model, opts, tried)
+			}
+		}
+		m.mu.RUnlock()
+	}
 	executor, okExecutor := m.Executor(provider)
 	if !okExecutor {
 		return nil, nil, &Error{Code: "executor_not_found", Message: "executor not registered"}
@@ -2381,7 +2539,7 @@ func (m *Manager) pickNextMixedLegacy(ctx context.Context, providers []string, m
 		if _, ok := m.executors[providerKey]; !ok {
 			continue
 		}
-		if modelKey != "" && registryRef != nil && !registryRef.ClientSupportsModel(candidate.ID, modelKey) {
+		if modelKey != "" && !m.authSupportsRouteModel(registryRef, candidate, model) {
 			continue
 		}
 		candidates = append(candidates, candidate)
@@ -2390,7 +2548,12 @@ func (m *Manager) pickNextMixedLegacy(ctx context.Context, providers []string, m
 		m.mu.RUnlock()
 		return nil, nil, "", &Error{Code: "auth_not_found", Message: "no auth available"}
 	}
-	selected, errPick := m.selector.Pick(ctx, "mixed", model, opts, candidates)
+	available, errAvailable := m.availableAuthsForRouteModel(candidates, "mixed", model, time.Now())
+	if errAvailable != nil {
+		m.mu.RUnlock()
+		return nil, nil, "", errAvailable
+	}
+	selected, errPick := m.selector.Pick(ctx, "mixed", selectionArgForSelector(m.selector, model), opts, available)
 	if errPick != nil {
 		m.mu.RUnlock()
 		return nil, nil, "", errPick
@@ -2442,6 +2605,29 @@ func (m *Manager) pickNextMixed(ctx context.Context, providers []string, model s
 	if len(eligibleProviders) == 0 {
 		return nil, nil, "", &Error{Code: "auth_not_found", Message: "no auth available"}
 	}
+	if strings.TrimSpace(model) != "" {
+		providerSet := make(map[string]struct{}, len(eligibleProviders))
+		for _, providerKey := range eligibleProviders {
+			providerSet[providerKey] = struct{}{}
+		}
+		m.mu.RLock()
+		for _, candidate := range m.auths {
+			if candidate == nil || candidate.Disabled {
+				continue
+			}
+			if _, ok := providerSet[strings.TrimSpace(strings.ToLower(candidate.Provider))]; !ok {
+				continue
+			}
+			if _, used := tried[candidate.ID]; used {
+				continue
+			}
+			if m.routeAwareSelectionRequired(candidate, model) {
+				m.mu.RUnlock()
+				return m.pickNextMixedLegacy(ctx, providers, model, opts, tried)
+			}
+		}
+		m.mu.RUnlock()
+	}
 
 	selected, providerKey, errPick := m.scheduler.pickMixed(ctx, eligibleProviders, model, opts, tried)
 	if errPick != nil && model != "" && shouldRetrySchedulerPick(errPick) {

sdk/cliproxy/auth/conductor_oauth_alias_suspension_test.go

@@ -0,0 +1,111 @@
+package auth
+
+import (
+	"context"
+	"net/http"
+	"sync"
+	"testing"
+	"time"
+
+	internalconfig "github.com/router-for-me/CLIProxyAPI/v6/internal/config"
+	"github.com/router-for-me/CLIProxyAPI/v6/internal/registry"
+	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
+)
+
+type aliasRoutingExecutor struct {
+	id string
+
+	mu            sync.Mutex
+	executeModels []string
+}
+
+func (e *aliasRoutingExecutor) Identifier() string { return e.id }
+
+func (e *aliasRoutingExecutor) Execute(_ context.Context, _ *Auth, req cliproxyexecutor.Request, _ cliproxyexecutor.Options) (cliproxyexecutor.Response, error) {
+	e.mu.Lock()
+	e.executeModels = append(e.executeModels, req.Model)
+	e.mu.Unlock()
+	return cliproxyexecutor.Response{Payload: []byte(req.Model)}, nil
+}
+
+func (e *aliasRoutingExecutor) ExecuteStream(context.Context, *Auth, cliproxyexecutor.Request, cliproxyexecutor.Options) (*cliproxyexecutor.StreamResult, error) {
+	return nil, &Error{HTTPStatus: http.StatusNotImplemented, Message: "ExecuteStream not implemented"}
+}
+
+func (e *aliasRoutingExecutor) Refresh(_ context.Context, auth *Auth) (*Auth, error) {
+	return auth, nil
+}
+
+func (e *aliasRoutingExecutor) CountTokens(context.Context, *Auth, cliproxyexecutor.Request, cliproxyexecutor.Options) (cliproxyexecutor.Response, error) {
+	return cliproxyexecutor.Response{}, &Error{HTTPStatus: http.StatusNotImplemented, Message: "CountTokens not implemented"}
+}
+
+func (e *aliasRoutingExecutor) HttpRequest(context.Context, *Auth, *http.Request) (*http.Response, error) {
+	return nil, &Error{HTTPStatus: http.StatusNotImplemented, Message: "HttpRequest not implemented"}
+}
+
+func (e *aliasRoutingExecutor) ExecuteModels() []string {
+	e.mu.Lock()
+	defer e.mu.Unlock()
+	out := make([]string, len(e.executeModels))
+	copy(out, e.executeModels)
+	return out
+}
+
+func TestManagerExecute_OAuthAliasBypassesBlockedRouteModel(t *testing.T) {
+	const (
+		provider    = "antigravity"
+		routeModel  = "claude-opus-4-6"
+		targetModel = "claude-opus-4-6-thinking"
+	)
+
+	manager := NewManager(nil, nil, nil)
+	executor := &aliasRoutingExecutor{id: provider}
+	manager.RegisterExecutor(executor)
+	manager.SetOAuthModelAlias(map[string][]internalconfig.OAuthModelAlias{
+		provider: {{
+			Name:  targetModel,
+			Alias: routeModel,
+			Fork:  true,
+		}},
+	})
+
+	auth := &Auth{
+		ID:       "oauth-alias-auth",
+		Provider: provider,
+		Status:   StatusActive,
+		ModelStates: map[string]*ModelState{
+			routeModel: {
+				Unavailable:    true,
+				Status:         StatusError,
+				NextRetryAfter: time.Now().Add(1 * time.Hour),
+			},
+		},
+	}
+	if _, errRegister := manager.Register(context.Background(), auth); errRegister != nil {
+		t.Fatalf("register auth: %v", errRegister)
+	}
+
+	reg := registry.GetGlobalRegistry()
+	reg.RegisterClient(auth.ID, provider, []*registry.ModelInfo{{ID: routeModel}, {ID: targetModel}})
+	t.Cleanup(func() {
+		reg.UnregisterClient(auth.ID)
+	})
+	manager.RefreshSchedulerEntry(auth.ID)
+
+	resp, errExecute := manager.Execute(context.Background(), []string{provider}, cliproxyexecutor.Request{Model: routeModel}, cliproxyexecutor.Options{})
+	if errExecute != nil {
+		t.Fatalf("execute error = %v, want success", errExecute)
+	}
+	if string(resp.Payload) != targetModel {
+		t.Fatalf("execute payload = %q, want %q", string(resp.Payload), targetModel)
+	}
+
+	gotModels := executor.ExecuteModels()
+	if len(gotModels) != 1 {
+		t.Fatalf("execute models len = %d, want 1", len(gotModels))
+	}
+	if gotModels[0] != targetModel {
+		t.Fatalf("execute model = %q, want %q", gotModels[0], targetModel)
+	}
+}

sdk/cliproxy/auth/conductor_overrides_test.go

@@ -12,6 +12,8 @@ import (
 	cliproxyexecutor "github.com/router-for-me/CLIProxyAPI/v6/sdk/cliproxy/executor"
 )
 
+const requestScopedNotFoundMessage = "Item with id 'rs_0b5f3eb6f51f175c0169ca74e4a85881998539920821603a74' not found. Items are not persisted when `store` is set to false. Try again with `store` set to true, or remove this item from your input."
+
 func TestManager_ShouldRetryAfterError_RespectsAuthRequestRetryOverride(t *testing.T) {
 	m := NewManager(nil, nil, nil)
 	m.SetRetryConfig(3, 30*time.Second, 0)
@@ -447,3 +449,114 @@ func TestManager_MarkResult_RespectsAuthDisableCoolingOverride(t *testing.T) {
 		t.Fatalf("expected NextRetryAfter to be zero when disable_cooling=true, got %v", state.NextRetryAfter)
 	}
 }
+
+func TestManager_MarkResult_RequestScopedNotFoundDoesNotCooldownAuth(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+
+	auth := &Auth{
+		ID:       "auth-1",
+		Provider: "openai",
+	}
+	if _, errRegister := m.Register(context.Background(), auth); errRegister != nil {
+		t.Fatalf("register auth: %v", errRegister)
+	}
+
+	model := "gpt-4.1"
+	m.MarkResult(context.Background(), Result{
+		AuthID:   auth.ID,
+		Provider: auth.Provider,
+		Model:    model,
+		Success:  false,
+		Error: &Error{
+			HTTPStatus: http.StatusNotFound,
+			Message:    requestScopedNotFoundMessage,
+		},
+	})
+
+	updated, ok := m.GetByID(auth.ID)
+	if !ok || updated == nil {
+		t.Fatalf("expected auth to be present")
+	}
+	if updated.Unavailable {
+		t.Fatalf("expected request-scoped 404 to keep auth available")
+	}
+	if !updated.NextRetryAfter.IsZero() {
+		t.Fatalf("expected request-scoped 404 to keep auth cooldown unset, got %v", updated.NextRetryAfter)
+	}
+	if state := updated.ModelStates[model]; state != nil {
+		t.Fatalf("expected request-scoped 404 to avoid model cooldown state, got %#v", state)
+	}
+}
+
+func TestManager_RequestScopedNotFoundStopsRetryWithoutSuspendingAuth(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+	executor := &authFallbackExecutor{
+		id: "openai",
+		executeErrors: map[string]error{
+			"aa-bad-auth": &Error{
+				HTTPStatus: http.StatusNotFound,
+				Message:    requestScopedNotFoundMessage,
+			},
+		},
+	}
+	m.RegisterExecutor(executor)
+
+	model := "gpt-4.1"
+	badAuth := &Auth{ID: "aa-bad-auth", Provider: "openai"}
+	goodAuth := &Auth{ID: "bb-good-auth", Provider: "openai"}
+
+	reg := registry.GetGlobalRegistry()
+	reg.RegisterClient(badAuth.ID, "openai", []*registry.ModelInfo{{ID: model}})
+	reg.RegisterClient(goodAuth.ID, "openai", []*registry.ModelInfo{{ID: model}})
+	t.Cleanup(func() {
+		reg.UnregisterClient(badAuth.ID)
+		reg.UnregisterClient(goodAuth.ID)
+	})
+
+	if _, errRegister := m.Register(context.Background(), badAuth); errRegister != nil {
+		t.Fatalf("register bad auth: %v", errRegister)
+	}
+	if _, errRegister := m.Register(context.Background(), goodAuth); errRegister != nil {
+		t.Fatalf("register good auth: %v", errRegister)
+	}
+
+	_, errExecute := m.Execute(context.Background(), []string{"openai"}, cliproxyexecutor.Request{Model: model}, cliproxyexecutor.Options{})
+	if errExecute == nil {
+		t.Fatal("expected request-scoped not-found error")
+	}
+	errResult, ok := errExecute.(*Error)
+	if !ok {
+		t.Fatalf("expected *Error, got %T", errExecute)
+	}
+	if errResult.HTTPStatus != http.StatusNotFound {
+		t.Fatalf("status = %d, want %d", errResult.HTTPStatus, http.StatusNotFound)
+	}
+	if errResult.Message != requestScopedNotFoundMessage {
+		t.Fatalf("message = %q, want %q", errResult.Message, requestScopedNotFoundMessage)
+	}
+
+	got := executor.ExecuteCalls()
+	want := []string{badAuth.ID}
+	if len(got) != len(want) {
+		t.Fatalf("execute calls = %v, want %v", got, want)
+	}
+	for i := range want {
+		if got[i] != want[i] {
+			t.Fatalf("execute call %d auth = %q, want %q", i, got[i], want[i])
+		}
+	}
+
+	updatedBad, ok := m.GetByID(badAuth.ID)
+	if !ok || updatedBad == nil {
+		t.Fatalf("expected bad auth to remain registered")
+	}
+	if updatedBad.Unavailable {
+		t.Fatalf("expected request-scoped 404 to keep bad auth available")
+	}
+	if !updatedBad.NextRetryAfter.IsZero() {
+		t.Fatalf("expected request-scoped 404 to keep bad auth cooldown unset, got %v", updatedBad.NextRetryAfter)
+	}
+	if state := updatedBad.ModelStates[model]; state != nil {
+		t.Fatalf("expected request-scoped 404 to avoid bad auth model cooldown state, got %#v", state)
+	}
+}

sdk/cliproxy/auth/conductor_update_test.go

@@ -47,3 +47,158 @@ func TestManager_Update_PreservesModelStates(t *testing.T) {
 		t.Fatalf("expected BackoffLevel to be %d, got %d", backoffLevel, state.Quota.BackoffLevel)
 	}
 }
+
+func TestManager_Update_DisabledExistingDoesNotInheritModelStates(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+
+	// Register a disabled auth with existing ModelStates.
+	if _, err := m.Register(context.Background(), &Auth{
+		ID:       "auth-disabled",
+		Provider: "claude",
+		Disabled: true,
+		Status:   StatusDisabled,
+		ModelStates: map[string]*ModelState{
+			"stale-model": {
+				Quota: QuotaState{BackoffLevel: 5},
+			},
+		},
+	}); err != nil {
+		t.Fatalf("register auth: %v", err)
+	}
+
+	// Update with empty ModelStates — should NOT inherit stale states.
+	if _, err := m.Update(context.Background(), &Auth{
+		ID:       "auth-disabled",
+		Provider: "claude",
+		Disabled: true,
+		Status:   StatusDisabled,
+	}); err != nil {
+		t.Fatalf("update auth: %v", err)
+	}
+
+	updated, ok := m.GetByID("auth-disabled")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth to be present")
+	}
+	if len(updated.ModelStates) != 0 {
+		t.Fatalf("expected disabled auth NOT to inherit ModelStates, got %d entries", len(updated.ModelStates))
+	}
+}
+
+func TestManager_Update_ActiveToDisabledDoesNotInheritModelStates(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+
+	// Register an active auth with ModelStates (simulates existing live auth).
+	if _, err := m.Register(context.Background(), &Auth{
+		ID:       "auth-a2d",
+		Provider: "claude",
+		Status:   StatusActive,
+		ModelStates: map[string]*ModelState{
+			"stale-model": {
+				Quota: QuotaState{BackoffLevel: 9},
+			},
+		},
+	}); err != nil {
+		t.Fatalf("register auth: %v", err)
+	}
+
+	// File watcher deletes config → synthesizes Disabled=true auth → Update.
+	// Even though existing is active, incoming auth is disabled → skip inheritance.
+	if _, err := m.Update(context.Background(), &Auth{
+		ID:       "auth-a2d",
+		Provider: "claude",
+		Disabled: true,
+		Status:   StatusDisabled,
+	}); err != nil {
+		t.Fatalf("update auth: %v", err)
+	}
+
+	updated, ok := m.GetByID("auth-a2d")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth to be present")
+	}
+	if len(updated.ModelStates) != 0 {
+		t.Fatalf("expected active→disabled transition NOT to inherit ModelStates, got %d entries", len(updated.ModelStates))
+	}
+}
+
+func TestManager_Update_DisabledToActiveDoesNotInheritStaleModelStates(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+
+	// Register a disabled auth with stale ModelStates.
+	if _, err := m.Register(context.Background(), &Auth{
+		ID:       "auth-d2a",
+		Provider: "claude",
+		Disabled: true,
+		Status:   StatusDisabled,
+		ModelStates: map[string]*ModelState{
+			"stale-model": {
+				Quota: QuotaState{BackoffLevel: 4},
+			},
+		},
+	}); err != nil {
+		t.Fatalf("register auth: %v", err)
+	}
+
+	// Re-enable: incoming auth is active, existing is disabled → skip inheritance.
+	if _, err := m.Update(context.Background(), &Auth{
+		ID:       "auth-d2a",
+		Provider: "claude",
+		Status:   StatusActive,
+	}); err != nil {
+		t.Fatalf("update auth: %v", err)
+	}
+
+	updated, ok := m.GetByID("auth-d2a")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth to be present")
+	}
+	if len(updated.ModelStates) != 0 {
+		t.Fatalf("expected disabled→active transition NOT to inherit stale ModelStates, got %d entries", len(updated.ModelStates))
+	}
+}
+
+func TestManager_Update_ActiveInheritsModelStates(t *testing.T) {
+	m := NewManager(nil, nil, nil)
+
+	model := "active-model"
+	backoffLevel := 3
+
+	// Register an active auth with ModelStates.
+	if _, err := m.Register(context.Background(), &Auth{
+		ID:       "auth-active",
+		Provider: "claude",
+		Status:   StatusActive,
+		ModelStates: map[string]*ModelState{
+			model: {
+				Quota: QuotaState{BackoffLevel: backoffLevel},
+			},
+		},
+	}); err != nil {
+		t.Fatalf("register auth: %v", err)
+	}
+
+	// Update with empty ModelStates — both sides active → SHOULD inherit.
+	if _, err := m.Update(context.Background(), &Auth{
+		ID:       "auth-active",
+		Provider: "claude",
+		Status:   StatusActive,
+	}); err != nil {
+		t.Fatalf("update auth: %v", err)
+	}
+
+	updated, ok := m.GetByID("auth-active")
+	if !ok || updated == nil {
+		t.Fatalf("expected auth to be present")
+	}
+	if len(updated.ModelStates) == 0 {
+		t.Fatalf("expected active auth to inherit ModelStates")
+	}
+	state := updated.ModelStates[model]
+	if state == nil {
+		t.Fatalf("expected model state to be present")
+	}
+	if state.Quota.BackoffLevel != backoffLevel {
+		t.Fatalf("expected BackoffLevel to be %d, got %d", backoffLevel, state.Quota.BackoffLevel)
+	}
+}

sdk/cliproxy/auth/custom_headers.go

@@ -0,0 +1,68 @@
+package auth
+
+import "strings"
+
+func ExtractCustomHeadersFromMetadata(metadata map[string]any) map[string]string {
+	if len(metadata) == 0 {
+		return nil
+	}
+	raw, ok := metadata["headers"]
+	if !ok || raw == nil {
+		return nil
+	}
+
+	out := make(map[string]string)
+	switch headers := raw.(type) {
+	case map[string]string:
+		for key, value := range headers {
+			name := strings.TrimSpace(key)
+			if name == "" {
+				continue
+			}
+			val := strings.TrimSpace(value)
+			if val == "" {
+				continue
+			}
+			out[name] = val
+		}
+	case map[string]any:
+		for key, value := range headers {
+			name := strings.TrimSpace(key)
+			if name == "" {
+				continue
+			}
+			rawVal, ok := value.(string)
+			if !ok {
+				continue
+			}
+			val := strings.TrimSpace(rawVal)
+			if val == "" {
+				continue
+			}
+			out[name] = val
+		}
+	default:
+		return nil
+	}
+
+	if len(out) == 0 {
+		return nil
+	}
+	return out
+}
+
+func ApplyCustomHeadersFromMetadata(auth *Auth) {
+	if auth == nil || len(auth.Metadata) == 0 {
+		return
+	}
+	headers := ExtractCustomHeadersFromMetadata(auth.Metadata)
+	if len(headers) == 0 {
+		return
+	}
+	if auth.Attributes == nil {
+		auth.Attributes = make(map[string]string)
+	}
+	for name, value := range headers {
+		auth.Attributes["header:"+name] = value
+	}
+}

sdk/cliproxy/auth/custom_headers_test.go

@@ -0,0 +1,50 @@
+package auth
+
+import (
+	"reflect"
+	"testing"
+)
+
+func TestExtractCustomHeadersFromMetadata(t *testing.T) {
+	meta := map[string]any{
+		"headers": map[string]any{
+			" X-Test ": " value ",
+			"":         "ignored",
+			"X-Empty":  "   ",
+			"X-Num":    float64(1),
+		},
+	}
+
+	got := ExtractCustomHeadersFromMetadata(meta)
+	want := map[string]string{"X-Test": "value"}
+	if !reflect.DeepEqual(got, want) {
+		t.Fatalf("ExtractCustomHeadersFromMetadata() = %#v, want %#v", got, want)
+	}
+}
+
+func TestApplyCustomHeadersFromMetadata(t *testing.T) {
+	auth := &Auth{
+		Metadata: map[string]any{
+			"headers": map[string]string{
+				"X-Test":  "new",
+				"X-Empty": "   ",
+			},
+		},
+		Attributes: map[string]string{
+			"header:X-Test": "old",
+			"keep":          "1",
+		},
+	}
+
+	ApplyCustomHeadersFromMetadata(auth)
+
+	if got := auth.Attributes["header:X-Test"]; got != "new" {
+		t.Fatalf("header:X-Test = %q, want %q", got, "new")
+	}
+	if _, ok := auth.Attributes["header:X-Empty"]; ok {
+		t.Fatalf("expected header:X-Empty to be absent, got %#v", auth.Attributes["header:X-Empty"])
+	}
+	if got := auth.Attributes["keep"]; got != "1" {
+		t.Fatalf("keep = %q, want %q", got, "1")
+	}
+}

sdk/cliproxy/auth/scheduler.go

@@ -219,6 +219,19 @@ func (s *authScheduler) pickMixed(ctx context.Context, providers []string, model
 	if len(normalized) == 0 {
 		return nil, "", &Error{Code: "provider_not_found", Message: "no provider supplied"}
 	}
+	if len(normalized) == 1 {
+		// When a single provider is eligible, reuse pickSingle so provider-specific preferences
+		// (for example Codex websocket transport) are applied consistently.
+		providerKey := normalized[0]
+		picked, errPick := s.pickSingle(ctx, providerKey, model, opts, tried)
+		if errPick != nil {
+			return nil, "", errPick
+		}
+		if picked == nil {
+			return nil, "", &Error{Code: "auth_not_found", Message: "no auth available"}
+		}
+		return picked, providerKey, nil
+	}
 	pinnedAuthID := pinnedAuthIDFromMetadata(opts.Metadata)
 	modelKey := canonicalModelKey(model)
 
@@ -293,12 +306,46 @@ func (s *authScheduler) pickMixed(ctx context.Context, providers []string, model
 	}
 
 	cursorKey := strings.Join(normalized, ",") + ":" + modelKey
-	start := 0
-	if len(normalized) > 0 {
-		start = s.mixedCursors[cursorKey] % len(normalized)
+	weights := make([]int, len(normalized))
+	segmentStarts := make([]int, len(normalized))
+	segmentEnds := make([]int, len(normalized))
+	totalWeight := 0
+	for providerIndex, shard := range candidateShards {
+		segmentStarts[providerIndex] = totalWeight
+		if shard != nil {
+			weights[providerIndex] = shard.readyCountAtPriorityLocked(false, bestPriority)
+		}
+		totalWeight += weights[providerIndex]
+		segmentEnds[providerIndex] = totalWeight
 	}
+	if totalWeight == 0 {
+		return nil, "", s.mixedUnavailableErrorLocked(normalized, model, tried)
+	}
+
+	startSlot := s.mixedCursors[cursorKey] % totalWeight
+	startProviderIndex := -1
+	for providerIndex := range normalized {
+		if weights[providerIndex] == 0 {
+			continue
+		}
+		if startSlot < segmentEnds[providerIndex] {
+			startProviderIndex = providerIndex
+			break
+		}
+	}
+	if startProviderIndex < 0 {
+		return nil, "", s.mixedUnavailableErrorLocked(normalized, model, tried)
+	}
+
+	slot := startSlot
 	for offset := 0; offset < len(normalized); offset++ {
-		providerIndex := (start + offset) % len(normalized)
+		providerIndex := (startProviderIndex + offset) % len(normalized)
+		if weights[providerIndex] == 0 {
+			continue
+		}
+		if providerIndex != startProviderIndex {
+			slot = segmentStarts[providerIndex]
+		}
 		providerKey := normalized[providerIndex]
 		shard := candidateShards[providerIndex]
 		if shard == nil {
@@ -308,7 +355,7 @@ func (s *authScheduler) pickMixed(ctx context.Context, providers []string, model
 		if picked == nil {
 			continue
 		}
-		s.mixedCursors[cursorKey] = providerIndex + 1
+		s.mixedCursors[cursorKey] = slot + 1
 		return picked, providerKey, nil
 	}
 	return nil, "", s.mixedUnavailableErrorLocked(normalized, model, tried)
@@ -667,16 +714,25 @@ func (m *modelScheduler) highestReadyPriorityLocked(preferWebsocket bool, predic
 	if m == nil {
 		return 0, false
 	}
+	if preferWebsocket {
+		// When downstream is websocket and Codex supports websocket transport, prefer websocket-enabled
+		// credentials even if they are in a lower priority tier than HTTP-only credentials.
+		for _, priority := range m.priorityOrder {
+			bucket := m.readyByPriority[priority]
+			if bucket == nil {
+				continue
+			}
+			if bucket.ws.pickFirst(predicate) != nil {
+				return priority, true
+			}
+		}
+	}
 	for _, priority := range m.priorityOrder {
 		bucket := m.readyByPriority[priority]
 		if bucket == nil {
 			continue
 		}
-		view := &bucket.all
-		if preferWebsocket && len(bucket.ws.flat) > 0 {
-			view = &bucket.ws
-		}
-		if view.pickFirst(predicate) != nil {
+		if bucket.all.pickFirst(predicate) != nil {
 			return priority, true
 		}
 	}
@@ -694,7 +750,7 @@ func (m *modelScheduler) pickReadyAtPriorityLocked(preferWebsocket bool, priorit
 		return nil
 	}
 	view := &bucket.all
-	if preferWebsocket && len(bucket.ws.flat) > 0 {
+	if preferWebsocket && bucket.ws.pickFirst(predicate) != nil {
 		view = &bucket.ws
 	}
 	var picked *scheduledAuth
@@ -709,6 +765,20 @@ func (m *modelScheduler) pickReadyAtPriorityLocked(preferWebsocket bool, priorit
 	return picked.auth
 }
 
+func (m *modelScheduler) readyCountAtPriorityLocked(preferWebsocket bool, priority int) int {
+	if m == nil {
+		return 0
+	}
+	bucket := m.readyByPriority[priority]
+	if bucket == nil {
+		return 0
+	}
+	if preferWebsocket && len(bucket.ws.flat) > 0 {
+		return len(bucket.ws.flat)
+	}
+	return len(bucket.all.flat)
+}
+
 // unavailableErrorLocked returns the correct unavailable or cooldown error for the shard.
 func (m *modelScheduler) unavailableErrorLocked(provider, model string, predicate func(*scheduledAuth) bool) error {
 	now := time.Now()

sdk/cliproxy/auth/scheduler_test.go

@@ -208,7 +208,33 @@ func TestSchedulerPick_CodexWebsocketPrefersWebsocketEnabledSubset(t *testing.T)
 	}
 }
 
-func TestSchedulerPick_MixedProvidersUsesProviderRotationOverReadyCandidates(t *testing.T) {
+func TestSchedulerPick_CodexWebsocketPrefersWebsocketEnabledAcrossPriorities(t *testing.T) {
+	t.Parallel()
+
+	scheduler := newSchedulerForTest(
+		&RoundRobinSelector{},
+		&Auth{ID: "codex-http", Provider: "codex", Attributes: map[string]string{"priority": "10"}},
+		&Auth{ID: "codex-ws-a", Provider: "codex", Attributes: map[string]string{"priority": "0", "websockets": "true"}},
+		&Auth{ID: "codex-ws-b", Provider: "codex", Attributes: map[string]string{"priority": "0", "websockets": "true"}},
+	)
+
+	ctx := cliproxyexecutor.WithDownstreamWebsocket(context.Background())
+	want := []string{"codex-ws-a", "codex-ws-b", "codex-ws-a"}
+	for index, wantID := range want {
+		got, errPick := scheduler.pickSingle(ctx, "codex", "", cliproxyexecutor.Options{}, nil)
+		if errPick != nil {
+			t.Fatalf("pickSingle() #%d error = %v", index, errPick)
+		}
+		if got == nil {
+			t.Fatalf("pickSingle() #%d auth = nil", index)
+		}
+		if got.ID != wantID {
+			t.Fatalf("pickSingle() #%d auth.ID = %q, want %q", index, got.ID, wantID)
+		}
+	}
+}
+
+func TestSchedulerPick_MixedProvidersUsesWeightedProviderRotationOverReadyCandidates(t *testing.T) {
 	t.Parallel()
 
 	scheduler := newSchedulerForTest(
@@ -218,8 +244,8 @@ func TestSchedulerPick_MixedProvidersUsesProviderRotationOverReadyCandidates(t *
 		&Auth{ID: "claude-a", Provider: "claude"},
 	)
 
-	wantProviders := []string{"gemini", "claude", "gemini", "claude"}
-	wantIDs := []string{"gemini-a", "claude-a", "gemini-b", "claude-a"}
+	wantProviders := []string{"gemini", "gemini", "claude", "gemini"}
+	wantIDs := []string{"gemini-a", "gemini-b", "claude-a", "gemini-a"}
 	for index := range wantProviders {
 		got, provider, errPick := scheduler.pickMixed(context.Background(), []string{"gemini", "claude"}, "", cliproxyexecutor.Options{}, nil)
 		if errPick != nil {
@@ -272,7 +298,7 @@ func TestSchedulerPick_MixedProvidersPrefersHighestPriorityTier(t *testing.T) {
 	}
 }
 
-func TestManager_PickNextMixed_UsesProviderRotationBeforeCredentialRotation(t *testing.T) {
+func TestManager_PickNextMixed_UsesWeightedProviderRotationBeforeCredentialRotation(t *testing.T) {
 	t.Parallel()
 
 	manager := NewManager(nil, &RoundRobinSelector{}, nil)
@@ -288,8 +314,8 @@ func TestManager_PickNextMixed_UsesProviderRotationBeforeCredentialRotation(t *t
 		t.Fatalf("Register(claude-a) error = %v", errRegister)
 	}
 
-	wantProviders := []string{"gemini", "claude", "gemini", "claude"}
-	wantIDs := []string{"gemini-a", "claude-a", "gemini-b", "claude-a"}
+	wantProviders := []string{"gemini", "gemini", "claude", "gemini"}
+	wantIDs := []string{"gemini-a", "gemini-b", "claude-a", "gemini-a"}
 	for index := range wantProviders {
 		got, _, provider, errPick := manager.pickNextMixed(context.Background(), []string{"gemini", "claude"}, "", cliproxyexecutor.Options{}, map[string]struct{}{})
 		if errPick != nil {
@@ -399,8 +425,8 @@ func TestManager_PickNextMixed_UsesSchedulerRotation(t *testing.T) {
 		t.Fatalf("Register(claude-a) error = %v", errRegister)
 	}
 
-	wantProviders := []string{"gemini", "claude", "gemini", "claude"}
-	wantIDs := []string{"gemini-a", "claude-a", "gemini-b", "claude-a"}
+	wantProviders := []string{"gemini", "gemini", "claude", "gemini"}
+	wantIDs := []string{"gemini-a", "gemini-b", "claude-a", "gemini-a"}
 	for index := range wantProviders {
 		got, _, provider, errPick := manager.pickNextMixed(context.Background(), []string{"gemini", "claude"}, "", cliproxyexecutor.Options{}, nil)
 		if errPick != nil {