ls-youtube-transcript

To get a timestamped markdown transcript from a YouTube video, this Skill fetches subtitle tracks via yt-dlp and parses caption segments into readable text with timestamps. It outputs a clean markdown transcript for quoting, note-taking, and searching.

Yes, you can extract auto-generated YouTube subtitles for archiving. The Skill uses yt-dlp to pull metadata and subtitle tracks, including auto-generated captions, and attempts to persist the transcript payloads into local databases for later retrieval.

Yes, you need yt-dlp to export YouTube captions into markdown because the Skill relies on it to fetch video metadata and JSON3 subtitle tracks. It then parses these caption segments into a timestamped markdown transcript.

The best way to search for specific quotes inside a YouTube video is to extract the full timestamped transcript into markdown. This converts the spoken content into searchable text, allowing you to find exact lines and their timestamps.

Extracting YouTube subtitles works for videos with only auto-captions enabled. The Skill retrieves auto-generated subtitle tracks through yt-dlp when manual captions are unavailable, parsing them into timestamped markdown lines.

A limitation when parsing YouTube JSON3 subtitle tracks is that successful extraction depends on yt-dlp fetching valid caption data. Additionally, archiving transcripts into Prismatica and LiteSuite storage is handled on a best-effort basis.