Voice Control

Manage voice channel settings and test audio hardware.

When to use

When user says start listening, stop listening, test microphone, test speaker, voice settings, or audio check.

How to use

Voice channel is managed via CLI commands (run_shell):

• Start voice: run_shell "voice_start"
• Stop voice: run_shell "voice_stop"
• Test TTS: run_shell "voice_test_tts "
• Test ASR: run_shell "voice_test_asr <pcm_file>"
• Check TTS backend: run_shell "set_voice_tts"
• Check ASR backend: run_shell "set_voice_asr"

Diagnostics

If voice isn't working:

1. run_shell "voice_test_tts hello" — test TTS output
2. Check if volc credentials are set: read_file /data/agent/config/config.json
3. Look for volc_appkey, volc_token, volc_api_key entries
4. If missing: tell user to run set_volc_key and set_volc_asr commands

Example

User: "开始语音监听" → run_shell "voice_start" → "语音通道已启动，正在监听。"

User: "测试一下扬声器" → run_shell "voice_test_tts 你好" → "TTS 测试完成，你应该听到了'你好'。"