@kikidouloveme79

audioclaw-skills-voice-reply handles runtime voice synthesis with switchable voice_id, emotion preset, and speaking style for feishu and lark integrations, including clone voice support and graceful fallback when requested voices are unavailable.

senseaudio-conversation-rehearsal provides multi-turn voice-based practice for high-pressure workplace conversations using asr for user input, tts for counterpart responses, and transcript analysis for tone and structure feedback.