fleet coral
Version: OpenClaw 2026.3.13 (61d171a)
OS: macOS arm64 (Apple Silicon)
Node: v25.8.1
What works:
- /vc join — bot successfully joins the voice channel
- Bot TTS config is set (OpenAI, gpt-4o-mini-tts, onyx voice)
- STT configured via tools.media.audio with gpt-4o-mini-transcribe
What doesn't work:
-
Bot never receives/transcribes any audio — zero voice events in logs after joining
-
No STT transcription fires regardless of how long I speak or how long the silence gap is
Packages verified loading from /opt/homebrew/lib/node_modules/openclaw:
- @discordjs/voice ✅
- @discordjs/opus ✅
- sodium-native ✅
- opusscript ✅ (had to manually npm install
opusscript + @discordjs/opus + sodium-native into openclaw's node_modules — they weren't present)
Privileged intents enabled in Discord Developer Portal:
- Message Content Intent ✅
- Server Members Intent ✅
- Presence Intent ✅
Logs show: Bot connects, joins voice channel on /vc join, but no audio receive events appear at all. No errors related to voice decoding.