🎥 youtube-summarizer

Extract transcripts from YouTube videos and generate comprehensive, detailed summaries

Version: 1.2.0 Status: ✨ Zero-Config | 🌍 Universal Platforms: GitHub Copilot CLI, Claude Code

---

Overview

The youtube-summarizer skill automates the extraction of YouTube video transcripts and generates verbose, structured summaries using the STAR + R-I-S-E framework. Perfect for documenting educational content, lectures, tutorials, or any informational videos without rewatching them.

---

Features

• 🎯 Automatic transcript extraction using youtube-transcript-api
• ✅ Video validation - Checks if video is accessible and has transcripts
• 🌍 Multi-language support - Prefers Portuguese, falls back to English
• 📊 Comprehensive summaries - Prioritizes detail and completeness
• 📝 Structured output - Markdown with headers, sections, insights
• 🔍 Metadata included - Video title, channel, duration, URL
• ⚡ Error handling - Clear messages for all failure scenarios
• 🛠️ Dependency management - Offers to install requirements automatically
• 📊 Progress gauge - Visual processing tracker across all steps
• 💾 Flexible save options - Summary-only, summary+transcript, or transcript-only (NEW v1.2.0)

---

Quick Start

Triggers

Activate this skill with any of these phrases:

# English
copilot> summarize this video: https://www.youtube.com/watch?v=VIDEO_ID
copilot> summarize youtube video https://youtu.be/VIDEO_ID
copilot> extract youtube transcript https://youtube.com/watch?v=VIDEO_ID

# Portuguese (also supported)
copilot> resume este video: https://www.youtube.com/watch?v=VIDEO_ID

First-Time Setup

The skill will automatically check for dependencies and offer to install them:

⚠️  youtube-transcript-api not installed

Would you like me to install it now?
- [x] Yes - Install with pip
- [ ] No - I'll install manually

Select "Yes" and the skill handles installation automatically.

---

Use Cases

1. Educational Video Documentation

copilot> summarize this video: https://www.youtube.com/watch?v=abc123

Output:

• Comprehensive summary of lecture content
• Key concepts and terminology
• Examples and practical applications
• Resources mentioned in the video

2. Technical Tutorial Analysis

copilot> summarize youtube video https://youtu.be/xyz789

Output:

• Step-by-step breakdown of tutorial
• Code snippets and commands mentioned
• Best practices highlighted
• Troubleshooting tips documented

3. Conference Talk Reference

copilot> extract youtube transcript https://youtube.com/watch?v=def456

Output:

• Speaker insights and arguments
• Statistics and data points
• Case studies and examples
• Q&A session summary

4. Language Learning Content

copilot> summarize youtube video https://youtu.be/ghi789

Output:

• Vocabulary and expressions used
• Grammar points explained
• Cultural references
• Practice exercises mentioned

5. Research and Investigation

copilot> summarize youtube video https://www.youtube.com/watch?v=jkl012

Output:

• Research findings presented
• Methodology explained
• Results and conclusions
• Future work suggestions

---

Output Structure

Every summary follows this comprehensive structure:

# [Video Title]

**Canal:** [Channel Name]
**Duração:** [Duration]
**URL:** [Video URL]
**Data de Publicação:** [Date]

---

## 📊 Síntese Executiva
[High-level overview, 2-3 paragraphs]

---

## 📝 Resumo Detalhado
### [Topic 1]
[Detailed analysis with examples, data, quotes]

### [Topic 2]
[Continued breakdown...]

---

## 💡 Principais Insights
- **Insight 1:** [Explanation]
- **Insight 2:** [Explanation]

---

## 📚 Conceitos e Terminologia
- **Term 1:** [Definition]
- **Term 2:** [Definition]

---

## 🔗 Recursos Mencionados
- [Resource 1]
- [Resource 2]

---

## 📌 Conclusão
[Final synthesis and key takeaways]

---

Requirements

• Python 3.x (usually pre-installed on macOS/Linux)
• pip (Python package manager)
• youtube-transcript-api by Julien Depoix (installed automatically by the skill)

Manual Installation (Optional)

If you prefer to install dependencies manually:

pip install youtube-transcript-api

---

Supported URL Formats

The skill recognizes these YouTube URL formats:

• https://www.youtube.com/watch?v=VIDEO_ID
• https://youtube.com/watch?v=VIDEO_ID
• https://youtu.be/VIDEO_ID
• https://m.youtube.com/watch?v=VIDEO_ID

---

Limitations

Videos That Work

✅ Public videos with auto-generated captions
✅ Videos with manual subtitles/captions
✅ Videos with transcripts in any supported language

Videos That Don't Work

❌ Private or unlisted videos
❌ Videos with transcripts disabled
❌ Age-restricted videos (may require authentication)
❌ Videos without any captions/subtitles

---

Error Messages

No Transcript Available

❌ No transcript available for this video

This skill requires videos with auto-generated captions or manual subtitles.
Unfortunately, transcripts are not enabled for this video.

Solution: Try a different video that has captions enabled.

Invalid URL

❌ Invalid YouTube URL format

Expected format examples:
- https://www.youtube.com/watch?v=VIDEO_ID
- https://youtu.be/VIDEO_ID

Solution: Ensure you're providing a complete, valid YouTube URL.

Video Not Accessible

❌ Unable to access video

Possible reasons:
1. Video is private or unlisted
2. Video has been removed
3. Invalid video ID

Solution: Verify the URL and ensure the video is public.

---

FAQ

Q: How long does it take to generate a summary?

A: Depends on video length:

• Short videos (5-10 min): 30-60 seconds
• Medium videos (20-40 min): 1-2 minutes
• Long videos (60+ min): 2-5 minutes

Q: Can I summarize videos in languages other than English/Portuguese?

A: Yes! The skill attempts to extract transcripts in the video's original language. If unavailable, it falls back to English.

Q: Will this work with YouTube Music videos?

A: Only if the music video has captions/transcripts enabled. Most music videos don't have transcripts.

Q: Can I customize the summary length?

A: The skill prioritizes completeness by design (verbose summaries). If you need shorter summaries, you can ask the AI to condense the output afterward.

Q: Does this download the video?

A: No. Only the text transcript is extracted via YouTube's API. No video files are downloaded.

Q: Can I save the summary to a file?

A: Yes! After the summary is generated, the skill offers flexible save options:

• Summary only - Markdown file with structured summary
• Summary + transcript - Markdown file with summary and raw transcript appended
• Transcript only - Plain text file with raw transcript (NEW in v1.2.0)
• Display only - No files saved, summary shown in terminal

Files are saved as resumo-{VIDEO_ID}-{YYYY-MM-DD}.md (summary) or transcript-{VIDEO_ID}-{YYYY-MM-DD}.txt (transcript-only).

Q: When should I save just the transcript?

A: Use the transcript-only option when you:

• Need raw content for further analysis
• Want to process the text with other tools
• Prefer to create your own summary later
• Need the transcript for documentation or archival purposes

---

Installation

Global Installation (Recommended)

Install the skill globally to use it across all projects:

# Clone the repository
git clone https://github.com/ericgandrade/cli-ai-skills.git
cd cli-ai-skills

# Run the install script
./scripts/install-skills.sh $(pwd)

This creates symlinks in:

• ~/.copilot/skills/youtube-summarizer/ (GitHub Copilot CLI)
• ~/.claude/skills/youtube-summarizer/ (Claude Code)

Repository Installation

Add to a specific project:

# Copy skill to your project
cp -r cli-ai-skills/.github/skills/youtube-summarizer .github/skills/

---

Contributing

Found a bug or have a feature request? Contributions welcome!

1. Fork the repository
2. Create a feature branch: git checkout -b feature/youtube-enhancement
3. Commit changes: git commit -m "feat(youtube-summarizer): add feature X"
4. Push and create a Pull Request

---

License

MIT License - see LICENSE for details.

---

Acknowledgments

• youtube-transcript-api by Julien Depoix - Python library for extracting YouTube video transcripts
• Anthropic STAR/R-I-S-E frameworks - For structured summarization

---

Built with ❤️ by Eric Andrade

Version 1.1.0 | Last updated: February 2026