3 Commits

Author	SHA1	Message	Date
kunthawat	261b0f3e91	feat: wire Opus decode + VAD turn detection to firmware-compatible reply ... Device streams continuous 20ms Opus frames with no turn markers. The loop now decodes Opus, feeds the VAD, fires one turn on end-of-speech, and replies with the firmware JSON contract (tts.state / stt.text) + Opus audio. - app/audio/opus.py: RealOpus/Opus16kEncoder/PassthroughOpus; fix add_dll_directory to use absolute paths (WinError 87) - app/xiaozhi/websocket.py: DeviceLoop.run() = decode -> VAD -> run_turn -> reply - app/config.py: AudioConfig (opus kind + VAD thresholds) - config.yaml: audio.opus=real, VAD 5/8/3000 frames - tests/test_app.py: utterance = loud burst + silence; asserts stt + audio + tts stop - requirements.txt: opuslib>=3 21/21 tests; live handshake verified via zhi.moreminimore.com on venv python.	2026-10-03 20:35:38 +07:00
kunthawat	a582c1ff36	feat: firmware-compatible handshake (header auth, transport=websocket) + /ota onboarding endpoint ... - auth from Authorization/Device-Id headers (firmware v2.5.0), frame fallback kept - server hello carries transport=websocket + session_id (firmware ParseServerHello) - /ota GET+POST returns {websocket:{url,token,version}, server_time:{timestamp,timezone_offset}, firmware:{version}} - config.yaml: device registered by MAC, ota block added - 21 tests pass (added OTA + header-auth coverage)	2026-10-03 16:35:41 +07:00
Macky	b8950b38a4	feat: Phase 1 testable bridge skeleton (18 tests green) ... Xiaozhi WebSocket endpoint with handshake/auth, mock STT/TTS/Hermes backends, Thai chunker, barge-in queue, latency logger, GPU planner, voice-profile guard (reasoning=none, session_search only). CON-002: no GPU/audio libs loaded at import. MUST-NOT-001: all model names/tokens from config.	2026-10-03 11:45:31 +07:00