Tapestry

One command to transform any learning resource (YouTube video, article, or PDF) into clean content and an actionable implementation plan. The master orchestrator that weaves together content extraction and action planning.

Core Concept

"Weave learning into action."

Tapestry automatically:

1. Detects what type of content you're sharing (YouTube, article, PDF)
2. Extracts clean, readable text from the source
3. Generates a Ship-Learn-Next action plan with 5 concrete reps
4. Saves both extracted content and implementation plan
5. Prompts you to start shipping your first rep immediately

One Command, Full Workflow

Instead of:

# Manual workflow (5+ steps)
1. Download YouTube transcript
2. Clean up duplicate lines
3. Save to file
4. Read the content
5. Brainstorm action items
6. Create implementation plan
7. Force yourself to start

Just use:

tapestry https://youtube.com/watch?v=example

Or:

weave https://blog.example.com/article

How It Works

What You Get

Two files created:

1. Extracted Content

   • Clean, readable text
   • No ads or navigation
   • Ready for reference

2. Ship-Learn-Next Plan

   • 5 concrete implementation reps
   • Weekly shipping targets
   • Reflection prompts
   • Iteration guidance

Usage Examples

From YouTube tutorial:

tapestry https://youtube.com/watch?v=dQw4w9WgXcQ

From technical article:

weave https://martinfowler.com/articles/microservices.html

From research paper PDF:

tapestry https://arxiv.org/pdf/2103.00020.pdf

Multiple resources at once:

I want to learn about React Server Components.
Please tapestry these resources:
- https://youtube.com/watch?v=example1
- https://react.dev/blog/2023/03/22/react-labs-what-we-have-been-working-on-march-2023
- https://example.com/react-server-components.pdf

Complete Workflow Example

User: tapestry https://youtube.com/watch?v=tutorial-on-tdd

Tapestry executes:

1. Detects YouTube URL
2. Downloads transcript with yt-dlp
3. Cleans duplicate lines
4. Saves "Test-Driven Development Tutorial - Transcript.txt"
5. Analyzes content for actionable lessons
6. Generates Ship-Learn-Next plan with 5 reps
7. Saves "Ship-Learn-Next Plan - Test-Driven Development.md"
8. Shows both file locations
9. Asks "Ready to start Rep 1?"

Required Tools

Automatically installed (macOS with Homebrew):

• yt-dlp - YouTube transcript download
• pdftotext (part of poppler) - PDF conversion

Optional (for better accuracy):

• reader-cli - Article extraction (npm)
• trafilatura - Fallback article extraction (Python)
• whisper - Fallback transcription (Python)

Installation:

# Core tools (macOS)
brew install yt-dlp poppler

# Optional enhancements
npm install -g @mozilla/readability-cli
pip3 install trafilatura openai-whisper

Philosophy: Ship-Learn-Next

Integration with Other Skills

Best Practices

Choose resources wisely:

• ✅ Practical tutorials with concrete techniques
• ✅ Deep dives on specific tools or patterns
• ✅ Case studies with implementation details
• ❌ High-level overviews without actionable content
• ❌ Purely theoretical discussions

After extraction:

• Review the Ship-Learn-Next plan
• Start Rep 1 within 24 hours (while motivated)
• Set calendar reminder for weekly shipping
• Share progress publicly for accountability

Managing multiple resources:

• Tapestry works great with batches
• Extract all content first, then review plans
• Combine related resources into one mega-plan
• Prioritize by what you can ship fastest

Common Use Cases

For Developers:

• Conference talks → Production implementations
• Tutorial videos → Portfolio projects
• Technical articles → Blog posts explaining concepts
• Documentation → Working code examples

For Content Creators:

• Podcast episodes → Article series
• Online courses → Original tutorials
• Research papers → Educational content
• Expert interviews → Actionable guides

For Learners:

• Study materials → Teaching others
• Industry trends → Experimental projects
• Best practices → Team documentation
• Tool comparisons → Hands-on evaluations

Limitations

Tapestry cannot:

• Access paywalled or login-required content
• Process content requiring JavaScript rendering
• Extract from sites with aggressive anti-scraping
• Generate perfect plans for every learning style

Workarounds:

• Log in before providing URLs (if authorized)
• Save articles as PDFs and use PDF mode
• Use browser extensions to grab content manually
• Customize generated plans to fit your needs

Troubleshooting

YouTube transcript fails:

• Check internet connection
• Verify video has captions/subtitles
• Falls back to Whisper (slower but works on any video)

Article extraction misses content:

• Try both reader-cli and trafilatura
• Some sites block scrapers (save as PDF instead)
• JavaScript-heavy sites may need manual copy-paste

PDF conversion produces gibberish:

• Some PDFs are scanned images, not text
• Use OCR tools separately for image-based PDFs
• Try different PDF viewers to verify content

About This Skill

Master orchestration skill that detects content type (YouTube, article, PDF), extracts clean content, and automatically creates Ship-Learn-Next action plans.