Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Extracts plain text transcripts from YouTube videos using the Supadata API, with automatic fallback to yt-dlp when the API is unavailable. Works on videos with auto-generated captions, not just manually added ones. Output has timestamps stripped, ready for further processing.
The dual-method fallback (API then yt-dlp) succeeds on videos without manual captions, where single-method tools or YouTube's own export would fail.
When to use it
- Skimming a conference talk before deciding whether to watch it
- Pulling exact quotes from an interview video for an article
- Researching a topic by scanning transcripts from multiple videos
- Creating searchable text archives of video content
- Checking if a tutorial covers a specific topic before committing time to watch it
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: get me the transcript of https://youtu.be/abc123
- 1Sends the video URL to the Supadata API
- 2Falls back to running yt-dlp locally if the API call fails
- 3Strips timestamps from the raw output
- 4Returns clean plain text transcript
Full plain text transcript of the video, ready for summarization or analysis
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.