fix(tts): include feishu in want_opus platforms

gateway/platforms/feishu.py

@@ -4865,6 +4865,12 @@ def _resolve_outbound_file_routing(
     ) -> tuple[str, str]:
         ext = Path(file_path).suffix.lower()
 
+        # If the caller explicitly requests audio, honor it.  Feishu accepts
+        # non-Opus audio uploaded as file_type=opus without codec validation,
+        # so MP3/WAV/M4A all render as voice bubbles when routed this way.
+        if requested_message_type == "audio":
+            return "opus", "audio"
+
         if ext in _FEISHU_OPUS_UPLOAD_EXTENSIONS:
             return "opus", "audio"
 

tests/gateway/test_feishu.py

@@ -2491,6 +2491,58 @@ async def _direct(func, *args, **kwargs):
         self.assertEqual(captured["message_request"].request_body.msg_type, "audio")
         self.assertEqual(captured["message_request"].request_body.content, '{"file_key": "file_audio_123"}')
 
+    def test_send_voice_routes_mp3_as_opus(self):
+        """MP3 files should be uploaded as file_type=opus when sent via send_voice,
+        even though .mp3 is not in _FEISHU_OPUS_UPLOAD_EXTENSIONS.  Feishu accepts
+        non-Opus audio uploaded as opus without codec validation."""
+        from gateway.config import PlatformConfig
+        from gateway.platforms.feishu import FeishuAdapter
+
+        adapter = FeishuAdapter(PlatformConfig())
+        captured = {}
+
+        class _FileAPI:
+            def create(self, request):
+                captured["upload_request"] = request
+                return SimpleNamespace(
+                    success=lambda: True,
+                    data=SimpleNamespace(file_key="file_mp3_456"),
+                )
+
+        class _MessageAPI:
+            def create(self, request):
+                captured["message_request"] = request
+                return SimpleNamespace(
+                    success=lambda: True,
+                    data=SimpleNamespace(message_id="om_mp3_msg"),
+                )
+
+        adapter._client = SimpleNamespace(
+            im=SimpleNamespace(
+                v1=SimpleNamespace(
+                    file=_FileAPI(),
+                    message=_MessageAPI(),
+                )
+            )
+        )
+
+        async def _direct(func, *args, **kwargs):
+            return func(*args, **kwargs)
+
+        with tempfile.NamedTemporaryFile("wb", suffix=".mp3", delete=False) as tmp:
+            tmp.write(b"mp3")
+            audio_path = tmp.name
+
+        try:
+            with patch("gateway.platforms.feishu.asyncio.to_thread", side_effect=_direct):
+                result = asyncio.run(adapter.send_voice(chat_id="oc_chat", audio_path=audio_path))
+        finally:
+            os.unlink(audio_path)
+
+        self.assertTrue(result.success)
+        self.assertEqual(captured["upload_request"].request_body.file_type, "opus")
+        self.assertEqual(captured["message_request"].request_body.msg_type, "audio")
+
     @patch.dict(os.environ, {}, clear=True)
     def test_build_post_payload_extracts_title_and_links(self):
         from gateway.config import PlatformConfig

tests/tools/test_tts_opus_routing.py

@@ -68,3 +68,32 @@ def fake_convert(path: str) -> str:
     assert result["voice_compatible"] is True
     assert result["media_tag"] == f"[[audio_as_voice]]\nMEDIA:{opus}"
     convert.assert_called_once_with(str(out))
+
+
+def test_minimax_feishu_converts_to_opus_voice(tmp_path, monkeypatch):
+    out = tmp_path / "speech.mp3"
+    opus = tmp_path / "speech.ogg"
+
+    def fake_convert(path: str) -> str:
+        assert path == str(out)
+        opus.write_bytes(b"ogg")
+        return str(opus)
+
+    convert = Mock(side_effect=fake_convert)
+
+    monkeypatch.setenv("HERMES_SESSION_PLATFORM", "feishu")
+    monkeypatch.setattr(tts_tool, "_load_tts_config", lambda: {"provider": "minimax"})
+    monkeypatch.setattr(
+        tts_tool,
+        "_generate_minimax_tts",
+        lambda _text, _output, _cfg: (_ := Path(_output).write_bytes(b"mp3")) or _output,
+    )
+    monkeypatch.setattr(tts_tool, "_convert_to_opus", convert)
+
+    result = json.loads(tts_tool.text_to_speech_tool("hello", output_path=str(out)))
+
+    assert result["success"] is True
+    assert result["file_path"] == str(opus)
+    assert result["voice_compatible"] is True
+    assert result["media_tag"] == f"[[audio_as_voice]]\nMEDIA:{opus}"
+    convert.assert_called_once_with(str(out))

tools/tts_tool.py

@@ -2064,7 +2064,7 @@ def text_to_speech_tool(
     # and needs ffmpeg for conversion.
     from gateway.session_context import get_session_env
     platform = get_session_env("HERMES_SESSION_PLATFORM", "").lower()
-    want_opus = (platform == "telegram")
+    want_opus = platform in {"telegram", "feishu"}
 
     # Determine output path
     if output_path: