youtube-learn - Video Belief Archaeology and Multimodal Analysis
Extracts speaker worldviews and core beliefs from videos using multimodal frame analysis and transcripts.
Tags
Updated: 2026-09-20Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Download video audio and frames
- Extract transcripts from audio
- Research speaker background and profile
- Analyze speaker core worldviews
- Compare beliefs with knowledge base
- Generate structured local analysis files
Inputs
- Video URL
- Existing knowledge index file
Outputs
- Analysis markdown report
- Video transcript file
- Speaker profile document
- Worldview markdown files
- Extracted frame image files
- Updated index map file
Requirements
- ffmpeg
- yt-dlp Python package
- youtube-transcript-api Python package
- Python environment
- KYMA, GROQ, or GEMINI API key