OpenClaw Skill - YouTube Transcript + Summary

Production-ready OpenClaw skill for YouTube video transcription and summarization.

Features

• ✅ Reliable Transcript Fetching - Dual-method approach bypasses YouTube rate limiting
• ✅ Batch Processing - Process multiple channels in one run
• ✅ AI-Powered Summaries - Generate structured, insightful summaries
• ✅ Cron-Friendly - Built for automated daily runs
• ✅ JSON Output - Flexible integration with any agent or platform
• ✅ Filters Shorts - Skip videos under 5 minutes

Quick Start

# Install
cd ~/.openclaw/skills
git clone https://github.com/happynocode/openclaw-skill-youtube.git youtube-summarizer
cd youtube-summarizer
./setup.sh

# Test single video
./youtube-summarizer --url "https://www.youtube.com/watch?v=dQw4w9WgXcQ"

# Scan channel (last 24 hours)
./youtube-summarizer --channel "UCSHZKyawb77ixDdsGog4iWA" --hours 24

# Daily batch (for cron)
./youtube-summarizer --config channels.json --daily --output /tmp/youtube_summary.json

How It Works

Transcript Fetching

Uses dual-method approach to ensure reliability:

1. Primary: innertube ANDROID client + Cloudflare Workers proxy

   • Bypasses YouTube's cloud IP restrictions
   • Works reliably from VPS environments

2. Fallback: youtube-transcript-api

   • Automatic fallback if primary method fails

This ensures consistent transcript fetching even when YouTube implements rate limiting.

Summary Generation

Generates structured summaries using LLM API:

### 🎯 核心问题/创新点
- 一句话概括视频要解决什么问题
- 有什么新颖的观点或技术突破

### 💡 核心论点（详细展开，每点2-3句话）
1. **论点1**：详细解释，包含具体数据、案例或证据
2. **论点2**：...

### 🛠️ 实操步骤（如果有）
1. 第一步：具体怎么做
2. 第二步：...

### 💰 价值与应用
- 这个内容对谁有价值
- 如何应用到实际工作/生活中

Configuration

Create channels.json:

{
  "channels": [
    {
      "name": "Lex Fridman",
      "id": "UCSHZKyawb77ixDdsGog4iWA",
      "url": "https://www.youtube.com/@lexfridman"
    },
    {
      "name": "Y Combinator",
      "id": "UCcefcZRL2oaA_uBNeo5UOWg",
      "url": "https://www.youtube.com/@ycombinator"
    }
  ],
  "hours_lookback": 24,
  "min_duration_seconds": 300,
  "max_videos_per_channel": 5
}

Output Format

{
  "generated_at": "2026-02-14T11:17:00Z",
  "items": [
    {
      "video_id": "dQw4w9WgXcQ",
      "title": "Video Title",
      "url": "https://youtube.com/watch?v=...",
      "channel": "Channel Name",
      "duration": "15:30",
      "published": "20260214",
      "has_transcript": true,
      "summary": "# Markdown summary...",
      "metadata": {
        "view_count": 12345,
        "like_count": 678
      }
    }
  ],
  "stats": {
    "total_videos": 5,
    "with_transcript": 4,
    "without_transcript": 1
  }
}

OpenClaw Integration

Cron Job Example

schedule:
  kind: cron
  expr: "0 8 * * *"  # 8 AM daily
payload:
  kind: agentTurn
  message: |
    Run YouTube daily summary:
    
    1. Execute skill:
       youtube-summarizer --config channels.json --daily --output /tmp/summary.json
    
    2. Read output and process each video
    
    3. Send to Discord/Telegram
    
    4. Sync to Notion

Agent Processing

import json

# Read output
with open("/tmp/youtube_summary.json") as f:
    data = json.load(f)

# Process each video
for item in data["items"]:
    if item["has_transcript"]:
        # Send to messaging platform
        send_message(item["summary"])
        
        # Sync to Notion
        sync_to_notion(item)

Advanced Configuration

Cloudflare Workers Proxy (Optional)

For improved reliability, deploy a Cloudflare Workers proxy:

export default {
  async fetch(request) {
    const url = new URL(request.url);
    const targetUrl = url.searchParams.get('url');
    
    if (!targetUrl) {
      return new Response('Missing url parameter', { status: 400 });
    }
    
    const response = await fetch(targetUrl, {
      headers: {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
      }
    });
    
    return new Response(response.body, {
      status: response.status,
      headers: {
        'Access-Control-Allow-Origin': '*',
        'Content-Type': response.headers.get('Content-Type') || 'application/json'
      }
    });
  }
};

Deploy to Cloudflare Workers, then update scripts/summarize.py:

CF_PROXY_URL = 'https://your-proxy.workers.dev/?url='

Dependencies

• yt-dlp - Video metadata extraction
• youtube-transcript-api - Transcript fetching (fallback)
• innertube - YouTube API client (primary method)
• Python 3.9+

All dependencies are installed automatically by setup.sh.

Environment Variables

• OPENCLAW_GATEWAY_TOKEN - OpenClaw API token for LLM calls
• OPENAI_API_KEY - OpenAI API key (alternative)

Configure in ~/.openclaw/openclaw.json or set as environment variables.

Model Configuration

Edit scripts/summarize.py to use your preferred model:

"model": "gpt-4o-mini",  # Change to your model

Examples:

• gpt-4o
• claude-3-5-sonnet-20241022
• gemini-2.0-flash-exp

Troubleshooting

Transcript fetch fails

• Video may not have captions
• Check has_transcript: false in output
• Try a different video

Rate limiting

• Skill uses innertube + proxy to bypass
• If still failing, check proxy configuration
• Reduce max_videos_per_channel in config

yt-dlp errors

• Update yt-dlp: pip install -U yt-dlp
• Check video is publicly accessible

Use Cases

• Daily Content Digest - Monitor your favorite channels
• Research & Curation - Collect insights from multiple sources
• Content Monitoring - Track specific topics or creators
• Knowledge Base - Build a searchable archive of video summaries

Contributing

Contributions welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

License

MIT License - see LICENSE for details.

Acknowledgments

• OpenClaw - AI agent framework
• innertube - YouTube API client
• yt-dlp - YouTube metadata extraction
• youtube-transcript-api - Transcript fetching

Links

• OpenClaw Documentation
• OpenClaw Skills Hub
• Report Issues

---

Built with ❤️ for the OpenClaw community