Commit Graph

1440 Commits

Author SHA1 Message Date
RGBadmin
8d8f5970ee fix(api): fallback to Metadata for priority/note on uploaded auths
buildAuthFileEntry now falls back to reading priority/note from
auth.Metadata when Attributes lacks them. This covers auths registered
via UploadAuthFile which bypass the synthesizer and only populate
Metadata from the raw JSON.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-15 17:36:11 +08:00
RGBadmin
f90120f846 fix(api): propagate note to Gemini virtual auths and align priority parsing
- Read note from Attributes (consistent with priority) in buildAuthFileEntry,
  fixing missing note on Gemini multi-project virtual auth cards.
- Propagate note from primary to virtual auths in SynthesizeGeminiVirtualAuths,
  mirroring existing priority propagation.
- Sync note/priority writes to both Metadata and Attributes in PatchAuthFileFields,
  with refactored nil-check to reduce duplication (review feedback).
- Validate priority type in fallback disk-read path instead of coercing all values
  to 0 via gjson.Int(), aligning with the auth-manager code path.
- Add regression tests for note synthesis, virtual-auth note propagation, and
  end-to-end multi-project Gemini note inheritance.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-15 16:47:01 +08:00
RGBadmin
5b6342e6ac feat(api): expose priority and note fields in GET /auth-files list response
The list endpoint previously omitted priority and note, which are stored
inside each auth file's JSON content. This adds them to both the normal
(auth-manager) and fallback (disk-read) code paths, and extends
PATCH /auth-files/fields to support writing the note field.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-14 14:47:31 +08:00
hkfires
c3d5dbe96f feat(model_registry): enhance model registration and refresh mechanisms 2026-03-13 10:56:39 +08:00
Luis Pater
817cebb321 Merge pull request #2082 from router-for-me/antigravity
Refactor Antigravity model handling and improve logging
2026-03-12 10:39:13 +08:00
Luis Pater
683f3709d6 Merge pull request #2076 from aikins01/fix/backfill-empty-function-response-names
fix: backfill empty functionResponse.name from preceding functionCall
2026-03-12 10:35:44 +08:00
hkfires
dbd42a42b2 fix(model_updater): clarify log message for model refresh failure 2026-03-12 10:32:04 +08:00
hkfires
dea3e74d35 feat(antigravity): refactor model handling and remove unused code 2026-03-12 09:24:45 +08:00
Aikins Laryea
a6c3042e34 refactor: remove redundant bounds checks per code review 2026-03-12 00:12:43 +00:00
Aikins Laryea
861537c9bd fix: backfill empty functionResponse.name from preceding functionCall
when Amp or Claude Code sends functionResponse with an empty name in Gemini
conversation history, the Gemini API rejects the request with 400
"Name cannot be empty". this fix backfills empty names from the
corresponding preceding functionCall parts using positional matching.

covers all three Gemini translator paths:
- gemini/gemini (direct API key)
- antigravity/gemini (OAuth)
- gemini-cli/gemini (Gemini CLI)

also switches fixCLIToolResponse pending group matching from LIFO to
FIFO to correctly handle multiple sequential tool call groups.

fixes #1903
2026-03-12 00:00:38 +00:00
Luis Pater
89d7be9525 Merge branch 'dev' into codex/custom-useragent-request 2026-03-11 22:55:50 +08:00
lang-911
70988d387b Add Codex websocket header defaults 2026-03-11 00:34:57 -07:00
Luis Pater
ddaa9d2436 Fixed: #2034
feat(proxy): centralize proxy handling with `proxyutil` package and enhance test coverage

- Added `proxyutil` package to simplify proxy handling across the codebase.
- Refactored various components (`executor`, `cliproxy`, `auth`, etc.) to use `proxyutil` for consistent and reusable proxy logic.
- Introduced support for "direct" proxy mode to explicitly bypass all proxies.
- Updated tests to validate proxy behavior (e.g., `direct`, HTTP/HTTPS, and SOCKS5).
- Enhanced YAML configuration documentation for proxy options.
2026-03-11 11:08:02 +08:00
Luis Pater
7b7b258c38 Fixed: #2022
test(translator): add tests for handling Claude system messages as string and array
2026-03-11 10:47:33 +08:00
hkfires
e333fbea3d feat(updater): update StartModelsUpdater to block until models refresh completes 2026-03-10 14:41:58 +08:00
hkfires
efbe36d1d4 feat(updater): change models refresh to one-time fetch on startup 2026-03-10 14:18:54 +08:00
hkfires
30d5c95b26 feat(registry): refresh model catalog from network 2026-03-10 14:02:54 +08:00
hkfires
d1e3195e6f feat(codex): register models by plan tier 2026-03-10 11:20:37 +08:00
Dominic Robinson
a1e0fa0f39 test(executor): cover string system prompt handling in checkSystemInstructionsWithMode 2026-03-09 12:40:27 +00:00
Dominic Robinson
5c9997cdac fix: Preserve system prompt when sent as a string instead of content block array 2026-03-09 07:38:11 +00:00
Kirill Turanskiy
338321e553 fix: use camelCase systemInstruction in OpenAI-to-Gemini translators
The Gemini v1internal (cloudcode-pa) and Antigravity Manager endpoints
require camelCase "systemInstruction" in request JSON. The current
snake_case "system_instruction" causes system prompts to be silently
ignored when routing through these endpoints.

Replace all "system_instruction" JSON keys with "systemInstruction" in
chat-completions and responses request translators.
2026-03-08 15:59:13 +03:00
Luis Pater
4f48e5254a Merge pull request #1957 from router-for-me/thinking
fix(translator): pass through adaptive thinking effort
2026-03-08 20:46:58 +08:00
hkfires
424711b718 fix(executor): use aiplatform base url for vertex api key calls 2026-03-08 20:13:12 +08:00
Luis Pater
b9153719b0 Merge pull request #1925 from shenshuoyaoyouguang/pr/openai-compat-pool-thinking
fix(openai-compat): improve pool fallback and preserve adaptive thinking
2026-03-08 01:05:05 +08:00
Luis Pater
631e5c8331 Merge pull request #1922 from shenshuoyaoyouguang/pr/model-registry-safety
fix(registry): clone model snapshots and invalidate available-model cache
2026-03-07 23:01:42 +08:00
Luis Pater
e9c60a0a67 Merge pull request #1910 from thebtf/fix/gemini-oauth-error-messages
fix: surface upstream error details in Gemini CLI OAuth onboarding UI
2026-03-07 22:25:18 +08:00
Luis Pater
98a1bb5a7f Merge pull request #1900 from rex-zsd/feature/add-gemini-3.1-flash-image-preview
feat(registry): add gemini-3.1-flash-image-preview model definition
2026-03-07 22:17:10 +08:00
Luis Pater
ca90487a8c Merge branch 'main' into feature/add-gemini-3.1-flash-image-preview 2026-03-07 22:16:09 +08:00
Luis Pater
1042489f85 Merge pull request #1893 from thebtf/fix/normalize-ttl-byte-preservation-mainline
fix: preserve original JSON bytes in normalizeCacheControlTTL
2026-03-07 22:14:13 +08:00
Luis Pater
38277c1ea6 Merge pull request #1875 from woqiqishi/fix/tool-use-id-sanitize
fix: sanitize tool_use.id to comply with Claude API regex ^[a-zA-Z0-9_-]+$
2026-03-07 22:06:36 +08:00
chujian
3a18f6fcca fix(registry): clone slice fields in model map output 2026-03-07 18:53:56 +08:00
chujian
099e734a02 fix(registry): always clone available model snapshots 2026-03-07 18:40:02 +08:00
chujian
97ef633c57 fix(registry): address review feedback 2026-03-07 17:36:57 +08:00
chujian
dae8463ba1 fix(registry): clone model snapshots and invalidate available-model cache 2026-03-07 16:59:23 +08:00
chujian
7c1299922e fix(openai-compat): improve pool fallback and preserve adaptive thinking 2026-03-07 16:54:28 +08:00
Luis Pater
7e6bb8fdc5 Merge origin/dev into pr-1774-review and resolve watcher conflicts 2026-03-07 11:12:42 +08:00
Luis Pater
9cee8ef87b Merge pull request #1684 from alexey-yanchenko/fix/input-audio-from-openai-to-antigravity
fix: preserve input_audio content parts when proxying to Antigravity
2026-03-07 10:12:28 +08:00
Luis Pater
93fb841bcb Fixed: #1670
test(translator): add unit tests for OpenAI to Claude requests and tool result handling

- Introduced tests for converting OpenAI requests to Claude with text, base64 images, and URL images in tool results.
- Refactored `convertClaudeToolResultContent` and related functionality to properly handle raw content with images and text.
- Updated conversion logic to streamline image handling for both base64 and URL formats.
2026-03-07 09:25:22 +08:00
Luis Pater
5ebc58fab4 refactor(executor): remove legacy connCreateSent logic and standardize response.create usage for all websocket events
- Simplified connection logic by removing `connCreateSent` and related state handling.
- Updated `buildCodexWebsocketRequestBody` to always use `response.create`.
- Added unit tests to validate `response.create` behavior and beta header preservation.
- Dropped unsupported `response.append` and outdated `response.done` event types.
2026-03-07 09:07:23 +08:00
Frad LEE
a8cbc68c3e feat(registry): add gemini 3.1 flash lite preview
- Add model to GetGeminiModels()
- Add model to GetGeminiVertexModels()
- Add model to GetGeminiCLIModels()
- Add model to GetAIStudioModels()
- Add to AntigravityModelConfig with thinking levels
- Update gemini-3-flash-preview description

Registers the new lightweight Gemini model across all provider
endpoints for cost-effective high-volume usage scenarios.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-06 20:52:28 +08:00
Kirill Turanskiy
11a795a01c fix: surface upstream error details in Gemini CLI OAuth onboarding UI
SetOAuthSessionError previously sent generic messages to the management
panel (e.g. "Failed to complete Gemini CLI onboarding"), hiding the
actual error returned by Google APIs. The specific error was only
written to the server log via log.Errorf, which is often inaccessible
in headless/Docker deployments.

Include the upstream error in all 8 OAuth error paths so the
management panel shows actionable messages like "no Google Cloud
projects available for this account" instead of a generic failure.
2026-03-06 13:06:37 +03:00
Luis Pater
2695a99623 fix(translator): conditionally remove service_tier from OpenAI response processing 2026-03-06 11:07:22 +08:00
zhongnan.rex
242aecd924 feat(registry): add gemini-3.1-flash-image-preview model definition 2026-03-06 10:50:04 +08:00
hkfires
ce8cc1ba33 fix(translator): pass through adaptive thinking effort 2026-03-06 09:13:32 +08:00
Kirill Turanskiy
97fdd2e088 fix: preserve original JSON bytes in normalizeCacheControlTTL when no TTL change needed
normalizeCacheControlTTL unconditionally re-serializes the entire request
body through json.Unmarshal/json.Marshal even when no TTL normalization
is needed. Go's json.Marshal randomizes map key order and HTML-escapes
<, >, & characters (to \u003c, \u003e, \u0026), producing different raw
bytes on every call.

Anthropic's prompt caching uses byte-prefix matching, so any byte-level
difference causes a cache miss. This means the ~119K system prompt and
tools are re-processed on every request when routed through CPA.

The fix adds a bool return to normalizeTTLForBlock to indicate whether
it actually modified anything, and skips the marshal step in
normalizeCacheControlTTL when no blocks were changed.
2026-03-05 22:28:01 +03:00
Luis Pater
9397f7049f fix(registry): simplify GPT 5.4 model description in static data 2026-03-06 02:32:56 +08:00
Luis Pater
8822f20d17 feat(registry): add GPT 5.4 model definition to static data 2026-03-06 02:23:53 +08:00
Xu Hong
553d6f50ea fix: sanitize tool_use.id to comply with Claude API regex ^[a-zA-Z0-9_-]+$
Add util.SanitizeClaudeToolID() to replace non-conforming characters in
tool_use.id fields across all five response translators (gemini, codex,
openai, antigravity, gemini-cli).

Upstream tool names may contain dots or other special characters
(e.g. "fs.readFile") that violate Claude's ID validation regex.
The sanitizer replaces such characters with underscores and provides
a generated fallback for empty IDs.

Fixes #1872, Fixes #1849

Made-with: Cursor
2026-03-06 00:10:09 +08:00
Luis Pater
f0e5a5a367 test(watcher): add unit test for server update timer cancellation and immediate reload logic
- Add `TestTriggerServerUpdateCancelsPendingTimerOnImmediate` to verify proper handling of server update debounce and timer cancellation.
- Fix logic in `triggerServerUpdate` to prevent duplicate timers and ensure proper cleanup of pending state.
2026-03-05 23:48:50 +08:00
Luis Pater
f6dfea9357 Merge pull request #1874 from constansino/fix/watcher-auth-event-storm-debounce
fix(watcher): 合并 auth 事件风暴下的回调触发,降低高 CPU
2026-03-05 23:29:56 +08:00