From d0c6353999fe6645ec2995f74202cf6ab399d6d6 Mon Sep 17 00:00:00 2001
From: seamusmore <35286167+seamusmore@users.noreply.github.com>
Date: Tue, 14 Jul 2026 18:25:34 +0800
Subject: [PATCH] fix(feishu): voice-note duration on upload, turn-scoped TTS
 dedup, audio path dedup

Trimmed cherry-pick of PR #40592 (duration + dedup hunks only; the
voice-classification hunk duplicates #29235 and the send_voice Opus
rewrite is out of scope for this inbound-focused PR):

- adapter.py: ffprobe duration (off-loop) attached to Feishu voice
  uploads via _build_file_upload_body(duration=...) and the audio
  message payload (#16524, #8300)
- gateway/run.py: TTS dedup narrowed to the current turn;
  _enrich_message_with_transcription dedups repeated audio paths

Refs #40592 #16524 #8300
---
 gateway/run.py | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/gateway/run.py b/gateway/run.py
index 3d2d9fa554979..a3ecc49a964fc 100644
--- a/gateway/run.py
+++ b/gateway/run.py
@@ -15902,14 +15902,19 @@ class GatewayRunner(GatewayAuthorizationMixin, GatewayKanbanWatchersMixin, Gatew
             )
             return False
 
-        # Dedup: agent already called TTS tool
+        # Dedup: agent already called TTS tool in THIS turn only
+        last_user_idx = None
+        for i, msg in enumerate(reversed(agent_messages)):
+            if msg.get("role") == "user":
+                last_user_idx = len(agent_messages) - 1 - i; break
+        turn_messages = agent_messages[last_user_idx:] if last_user_idx is not None else agent_messages
         has_agent_tts = any(
             msg.get("role") == "assistant"
             and any(
                 (tc.get("function") or {}).get("name") == "text_to_speech"
                 for tc in (msg.get("tool_calls") or [])
             )
-            for msg in agent_messages
+            for msg in turn_messages
         )
         if has_agent_tts:
             return False
@@ -18058,6 +18063,8 @@ class GatewayRunner(GatewayAuthorizationMixin, GatewayKanbanWatchersMixin, Gatew
                 list if every clip failed or STT is disabled. Callers can use
                 this to echo transcripts back to the user before the agent loop.
         """
+        seen = set()
+        audio_paths = [p for p in audio_paths if p not in seen and not seen.add(p)]
         if not getattr(self.config, "stt_enabled", True):
             notes = []
             for path in audio_paths: