YouTube Video Extractor Skill

This skill extracts comprehensive data from YouTube videos and formats it into an Obsidian note using the provided template.

Usage

When given a YouTube URL, follow these steps:

1. •

   Extract Video Data: Run the extraction script to get metadata, transcript, and thumbnail.

   • •Command: python scripts/extract_youtube.py <YOUTUBE_URL>
   • •This will output JSON with extracted data.
2. •

   Generate Obsidian Note: Use the extracted data to fill the template from ../../../obsidian-templates/templates/youtube-video-notes/advanced.md

3. •

   Fill Template Placeholders:

   • •Replace {{date}} with current date (YYYY-MM-DD)
   • •Replace {{video_url}} with the YouTube URL
   • •Replace {{channel}} with channel name
   • •Replace {{title}} with video title
   • •Replace {{duration}} with video duration (HH:MM:SS)
   • •Replace {{topic}} with inferred topic (e.g., from title or description)
   • •Add transcript content to the note if available
   • •Fill in Key Takeaways, Quotes, etc. from transcript if possible (or leave as template)
4. •

   Output the Complete Note: Provide the filled template as the final output.

Dependencies

• •Python 3.8+
• •External libraries (install via pip install -r requirements.txt):
  • •yt-dlp: For video metadata
  • •youtube-transcript-api: For extracting transcripts
  • •google-generativeai: For AI analysis and summarization
  • •python-dotenv: For managing environment variables

Examples

Input: https://youtu.be/OPH4yNvuMr8

Output: A complete Obsidian note with filled frontmatter and sections.

Edge Cases

• •If no transcript available, note that in the template.
• •If video is private or unavailable, handle error gracefully.
• •For long transcripts, summarize or truncate appropriately.