fix(discord): sync realtime voice playback timestamps

CHANGELOG.md

@@ -10,6 +10,7 @@ Docs: https://docs.openclaw.ai
 
 ### Fixes
 
+- Discord/voice: synthesize realtime playback timestamps from emitted Discord PCM so OpenAI realtime barge-in truncation no longer sees `audioEndMs=0` and skips legitimate interruptions.
 - Plugin SDK: keep activated linked plugin runtime facades loadable when bundled plugin fallback is disabled. Thanks @shakkernerd.
 - Feishu: auto-thread `message(action="send")` replies inside the topic when the active session is group_topic or group_topic_sender, and propagate `replyInThread` through text, card, and media outbound adapters so topic-scoped sessions no longer post at the group root. Fixes #74903. (#77151) Thanks @ai-hpc.
 - WhatsApp: pass routing context into voice-note transcript echo preflight so echoed transcripts can deliver to the originating chat. Fixes #79778. (#79788) Thanks @hclsys.

extensions/discord/src/voice/manager.e2e.test.ts

@@ -287,6 +287,7 @@ describe("DiscordVoiceManager", () => {
     realtimeSessionMock.sendAudio.mockClear();
     realtimeSessionMock.sendUserMessage.mockClear();
     realtimeSessionMock.handleBargeIn.mockClear();
+    realtimeSessionMock.setMediaTimestamp.mockClear();
     realtimeSessionMock.submitToolResult.mockClear();
     createRealtimeVoiceBridgeSessionMock.mockClear();
     createRealtimeVoiceBridgeSessionMock.mockReturnValue(realtimeSessionMock);
@@ -641,7 +642,12 @@ describe("DiscordVoiceManager", () => {
     bridgeParams?.audioSink?.sendAudio(Buffer.alloc(480));
     turn?.sendInputAudio(Buffer.alloc(3840));
 
+    expect(realtimeSessionMock.setMediaTimestamp).toHaveBeenCalledWith(0);
+    expect(realtimeSessionMock.setMediaTimestamp).toHaveBeenCalledWith(10);
     expect(realtimeSessionMock.handleBargeIn).toHaveBeenCalled();
+    const lastTimestampCall = realtimeSessionMock.setMediaTimestamp.mock.invocationCallOrder.at(-1);
+    const firstBargeInCall = realtimeSessionMock.handleBargeIn.mock.invocationCallOrder[0];
+    expect(lastTimestampCall).toBeLessThan(firstBargeInCall);
     expect(player.stop).not.toHaveBeenCalled();
     expect(realtimeSessionMock.sendAudio).toHaveBeenCalled();
   });

extensions/discord/src/voice/realtime.ts

@@ -43,6 +43,7 @@ const DISCORD_REALTIME_PENDING_SPEAKER_CONTEXT_LIMIT = 32;
 const DISCORD_REALTIME_LOG_PREVIEW_CHARS = 500;
 const DISCORD_REALTIME_DEFAULT_MIN_BARGE_IN_AUDIO_END_MS = 250;
 const DISCORD_REALTIME_FORCED_CONSULT_FALLBACK_DELAY_MS = 200;
+const REALTIME_PCM16_BYTES_PER_SAMPLE = 2;
 
 export type DiscordVoiceMode = "stt-tts" | "agent-proxy" | "bidi";
 
@@ -245,6 +246,7 @@ export class DiscordRealtimeVoiceSession implements VoiceRealtimeSession {
   private consultPolicy: "auto" | "always" = "auto";
   private pendingAgentProxyConsultContexts: PendingAgentProxyConsultContext[] = [];
   private readonly pendingSpeakerTurns: PendingSpeakerTurn[] = [];
+  private outputAudioTimestampMs = 0;
   private readonly playerIdleHandler = () => {
     this.resetOutputStream();
   };
@@ -430,6 +432,7 @@ export class DiscordRealtimeVoiceSession implements VoiceRealtimeSession {
     if (!this.isBargeInEnabled()) {
       return;
     }
+    this.syncOutputAudioTimestamp();
     this.bridge?.handleBargeIn({ audioPlaybackActive: true });
   }
 
@@ -450,8 +453,13 @@ export class DiscordRealtimeVoiceSession implements VoiceRealtimeSession {
     if (discordPcm.length === 0) {
       return;
     }
+    this.syncOutputAudioTimestamp();
     const stream = this.ensureOutputStream();
     stream.write(discordPcm);
+    this.outputAudioTimestampMs += pcm16MonoDurationMs(
+      realtimePcm24kMono,
+      REALTIME_VOICE_AUDIO_FORMAT_PCM16_24KHZ.sampleRateHz,
+    );
   }
 
   private ensureOutputStream(): PassThrough {
@@ -485,10 +493,15 @@ export class DiscordRealtimeVoiceSession implements VoiceRealtimeSession {
   private resetOutputStream(): void {
     const stream = this.outputStream;
     this.outputStream = null;
+    this.outputAudioTimestampMs = 0;
     stream?.end();
     stream?.destroy();
   }
 
+  private syncOutputAudioTimestamp(): void {
+    this.bridge?.setMediaTimestamp(Math.floor(this.outputAudioTimestampMs));
+  }
+
   private handleToolCall(
     event: RealtimeVoiceToolCallEvent,
     session: RealtimeVoiceBridgeSession,
@@ -640,6 +653,7 @@ export class DiscordRealtimeVoiceSession implements VoiceRealtimeSession {
     logger.info(
       `discord voice: realtime forced agent consult starting chars=${question.length} voiceSession=${this.params.entry.voiceSessionKey} supervisorSession=${this.params.entry.route.sessionKey} agent=${this.params.entry.route.agentId} speaker=${context.speakerLabel} owner=${context.senderIsOwner}`,
     );
+    this.syncOutputAudioTimestamp();
     this.bridge?.handleBargeIn({ audioPlaybackActive: true, force: true });
     this.clearOutputAudio();
     try {
@@ -718,6 +732,14 @@ function isDiscordRealtimeSpeakerContext(value: unknown): value is DiscordRealti
   );
 }
 
+function pcm16MonoDurationMs(audio: Buffer, sampleRate: number): number {
+  if (audio.length === 0 || sampleRate <= 0) {
+    return 0;
+  }
+  const samples = audio.length / REALTIME_PCM16_BYTES_PER_SAMPLE;
+  return (samples * 1000) / sampleRate;
+}
+
 function buildProviderConfigs(
   realtimeConfig: DiscordRealtimeVoiceConfig,
 ): Record<string, RealtimeVoiceProviderConfig | undefined> | undefined {