Realtime-Voice-Clone-Chinese/SKILL.md

name, description

name	description
MockingBird	AI voice cloning and TTS toolkit. The only active skill is `speak`.

MockingBird

Voice cloning / text-to-speech toolkit built on encoder-synthesizer-vocoder pipeline.

Available Skills

Skill	Path	Purpose
speak	skills/speak/SKILL.md	Text-to-speech (Kokoro local / Noiz cloud), timeline rendering, voice cloning

Project Layout

• models/ — encoder, synthesizer, vocoder, ppg2mel, ppg_extractor
• control/ — CLI and GUI entry points (toolbox, mkgui)
• skills/speak/ — TTS skill scripts (tts.sh, noiz_tts.py, render_timeline.py, text_to_srt.py)
• data/ckpt/ — pretrained checkpoints

When to Use

If the user asks for anything related to speech synthesis, TTS, voice cloning, dubbing, or audio generation, read and follow skills/speak/SKILL.md.