Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- config.example default provider change to elevenlabs
What This Skill Does
Transcribes audio files to timestamped lyrics using OpenAI Whisper or ElevenLabs Scribe. Outputs LRC, SRT, or JSON formats with word-level timestamps. Includes a post-transcription correction workflow to fix common misrecognition errors in sung lyrics.
Supports two transcription backends with word-level timestamps and includes a structured correction step that most raw transcription tools skip.
When to use it
- Generating LRC files for music video rendering
- Creating subtitles for song covers or demos
- Extracting timestamped lyrics from foreign-language tracks
- Syncing lyrics to audio for karaoke players
- Correcting AI-generated transcriptions against original lyrics
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: transcribe this Mandarin pop song to an LRC file
- 1Check if the active provider's API key is configured
- 2Run the transcription script with --audio and --language zh flags
- 3Read the output LRC file and the original lyrics reference
- 4Correct misrecognized words while preserving all timestamps
- 5Write the corrected LRC back to disk
A timestamped LRC file with accurate lyrics saved to acestep_output/<filename>.lrc
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.