youtube-learning-sync

name: youtube-learning-sync description: Ingest YouTube videos/watch-history items, extract transcripts, summarize learning, and produce daily/weekly learning briefs with actions. Use when Mat asks to process YouTube links/history, create learning summaries, compare ideas across videos, or build a recurring YouTube-learning workflow.

youtube-learning-sync

Use this skill to turn YouTube consumption into structured learning outputs.

Core workflow

1. Collect input

• Single URL(s), playlist URLs, or exported history files.

2. Extract transcript

• First try scripts/youtube_transcript_fetch.py.
• If captions are unavailable, use yt-dlp fallback to fetch metadata/captions where possible.
• If no captions exist, mark the item as transcript_unavailable and continue.

3. Summarize

• Produce: key ideas, notable claims, practical takeaways, open questions, and action items.
• Keep summaries decision-oriented.

4. Deliver briefs

• Daily brief: what was watched, what mattered, what to do next.
• Weekly brief: recurring themes, contradictions, changing beliefs, recommended experiments.

Commands

Transcript extraction:

python3 scripts/youtube_transcript_fetch.py <youtube-url-or-id>
python3 scripts/youtube_transcript_fetch.py <youtube-url-or-id> --json

Optional metadata/caption fallback:

yt-dlp --skip-download --write-auto-subs --sub-langs en --print "%(id)s\t%(title)s\t%(channel)s\t%(upload_date)s" <youtube-url>

Output format (default)

For each video:

• Title / URL
• 5 bullet summary
• 3 key takeaways
• 1 action for Mat
• Confidence (high/medium/low) based on transcript quality

Daily brief:

• What changed today
• Why it matters
• Recommended next action
• What HAL can do next

Notes

• Prefer transcript-backed summaries over title-only summaries.
• Flag weak evidence explicitly.
• Keep output concise and useful for decisions.