fix: harden discord audio preflight mention detection (#32136) (thanks @jnMetaCode)

CHANGELOG.md

@@ -62,6 +62,7 @@ Docs: https://docs.openclaw.ai
 - Gateway/Webchat NO_REPLY streaming: suppress assistant lead-fragment deltas that are prefixes of `NO_REPLY` and keep final-message buffering in sync, preventing partial `NO` leaks on silent-response runs while preserving legitimate short replies. (#32073) Thanks @liuxiaopai-ai.
 - Tools/fsPolicy propagation: honor `tools.fs.workspaceOnly` for image/pdf local-root allowlists so non-sandbox media paths outside workspace are rejected when workspace-only mode is enabled. (#31882) Thanks @justinhuangcode.
 - Daemon/Homebrew runtime pinning: resolve Homebrew Cellar Node paths to stable Homebrew-managed symlinks (including versioned formulas like `node@22`) so gateway installs keep the intended runtime across brew upgrades. (#32185) Thanks @scoootscooob.
+- Discord/audio preflight mentions: detect audio attachments via Discord `content_type` and gate preflight transcription on typed text (not media placeholders), so guild voice-note mentions are transcribed and matched correctly. (#32136) Thanks @jnMetaCode.
 - Memory/LanceDB embeddings: forward configured `embedding.dimensions` into OpenAI embeddings requests so vector size and API output dimensions stay aligned when dimensions are explicitly configured. (#32036) Thanks @scotthuang.
 - Failover/error classification: treat HTTP `529` (provider overloaded, common with Anthropic-compatible APIs) as `rate_limit` so model failover can engage instead of misclassifying the error path. (#31854) Thanks @bugkill3r.
 - Plugin command/runtime hardening: validate and normalize plugin command name/description at registration boundaries, and guard Telegram native menu normalization paths so malformed plugin command specs cannot crash startup (`trim` on undefined). (#31997) Fixes #31944. Thanks @liuxiaopai-ai.

src/discord/monitor/message-handler.preflight.test.ts

@@ -1,5 +1,11 @@
 import { ChannelType } from "@buape/carbon";
-import { beforeEach, describe, expect, it } from "vitest";
+import { beforeEach, describe, expect, it, vi } from "vitest";
+
+const transcribeFirstAudioMock = vi.hoisted(() => vi.fn());
+
+vi.mock("../../media-understanding/audio-preflight.js", () => ({
+  transcribeFirstAudio: (...args: unknown[]) => transcribeFirstAudioMock(...args),
+}));
 import {
   __testing as sessionBindingTesting,
   registerSessionBindingAdapter,
@@ -74,6 +80,7 @@ describe("resolvePreflightMentionRequirement", () => {
 describe("preflightDiscordMessage", () => {
   beforeEach(() => {
     sessionBindingTesting.resetSessionBindingAdaptersForTests();
+    transcribeFirstAudioMock.mockReset();
   });
 
   it("bypasses mention gating in bound threads for allowed bot senders", async () => {
@@ -165,6 +172,101 @@ describe("preflightDiscordMessage", () => {
     expect(result?.boundSessionKey).toBe(threadBinding.targetSessionKey);
     expect(result?.shouldRequireMention).toBe(false);
   });
+
+  it("uses attachment content_type for guild audio preflight mention detection", async () => {
+    transcribeFirstAudioMock.mockResolvedValue("hey openclaw");
+
+    const channelId = "channel-audio-1";
+    const client = {
+      fetchChannel: async (id: string) => {
+        if (id === channelId) {
+          return {
+            id: channelId,
+            type: ChannelType.GuildText,
+            name: "general",
+          };
+        }
+        return null;
+      },
+    } as unknown as import("@buape/carbon").Client;
+
+    const message = {
+      id: "m-audio-1",
+      content: "",
+      timestamp: new Date().toISOString(),
+      channelId,
+      attachments: [
+        {
+          id: "att-1",
+          url: "https://cdn.discordapp.com/attachments/voice.ogg",
+          content_type: "audio/ogg",
+          filename: "voice.ogg",
+        },
+      ],
+      mentionedUsers: [],
+      mentionedRoles: [],
+      mentionedEveryone: false,
+      author: {
+        id: "user-1",
+        bot: false,
+        username: "Alice",
+      },
+    } as unknown as import("@buape/carbon").Message;
+
+    const result = await preflightDiscordMessage({
+      cfg: {
+        session: {
+          mainKey: "main",
+          scope: "per-sender",
+        },
+        messages: {
+          groupChat: {
+            mentionPatterns: ["openclaw"],
+          },
+        },
+      } as import("../../config/config.js").OpenClawConfig,
+      discordConfig: {} as NonNullable<
+        import("../../config/config.js").OpenClawConfig["channels"]
+      >["discord"],
+      accountId: "default",
+      token: "token",
+      runtime: {} as import("../../runtime.js").RuntimeEnv,
+      botUserId: "openclaw-bot",
+      guildHistories: new Map(),
+      historyLimit: 0,
+      mediaMaxBytes: 1_000_000,
+      textLimit: 2_000,
+      replyToMode: "all",
+      dmEnabled: true,
+      groupDmEnabled: true,
+      ackReactionScope: "direct",
+      groupPolicy: "open",
+      threadBindings: createNoopThreadBindingManager("default"),
+      data: {
+        channel_id: channelId,
+        guild_id: "guild-1",
+        guild: {
+          id: "guild-1",
+          name: "Guild One",
+        },
+        author: message.author,
+        message,
+      } as unknown as import("./listeners.js").DiscordMessageEvent,
+      client,
+    });
+
+    expect(transcribeFirstAudioMock).toHaveBeenCalledTimes(1);
+    expect(transcribeFirstAudioMock).toHaveBeenCalledWith(
+      expect.objectContaining({
+        ctx: expect.objectContaining({
+          MediaUrls: ["https://cdn.discordapp.com/attachments/voice.ogg"],
+          MediaTypes: ["audio/ogg"],
+        }),
+      }),
+    );
+    expect(result).not.toBeNull();
+    expect(result?.wasMentioned).toBe(true);
+  });
 });
 
 describe("shouldIgnoreBoundThreadWebhookMessage", () => {

src/discord/monitor/message-handler.preflight.ts

@@ -504,11 +504,13 @@ export async function preflightDiscordMessage(
   const hasAudioAttachment = message.attachments?.some((att: { content_type?: string }) =>
     att.content_type?.startsWith("audio/"),
   );
+  const hasTypedText = Boolean(message.content?.trim());
   const needsPreflightTranscription =
     !isDirectMessage &&
     shouldRequireMention &&
     hasAudioAttachment &&
-    !baseText &&
+    // `baseText` includes media placeholders; gate on typed text only.
+    !hasTypedText &&
     mentionRegexes.length > 0;
 
   if (needsPreflightTranscription) {
@@ -541,10 +543,11 @@ export async function preflightDiscordMessage(
     }
   }
 
+  const mentionText = hasTypedText ? baseText : "";
   const wasMentioned =
     !isDirectMessage &&
     matchesMentionWithExplicit({
-      text: baseText,
+      text: mentionText,
       mentionRegexes,
       explicit: {
         hasAnyMention,