Video Sum

A Model Context Protocol (MCP) server that extracts content from multiple video platforms and generates intelligent knowledge graphs.

Features

🌐 Multi-Platform Support

• Douyin (TikTok China) - Short video content extraction
• Bilibili - Video and live streaming content
• Xiaohongshu (Little Red Book) - Social media posts with OCR support
• Zhihu - Q&A platform content

✨ Advanced Capabilities

• OCR Text Recognition - Extract text from images using PaddleOCR
• Knowledge Graph Generation - Intelligent content structuring
• Chinese Content Optimization - Specialized processing for Chinese text
• Context-Aware Extraction - Smart content understanding and quality control

Installation

Prerequisites

• Python 3.8 or higher
• Anaconda (recommended for dependency management)

Setup

1. Clone the repository:

git clone https://github.com/fakad/video-sum-mcp.git
cd video-sum-mcp

2. Create and activate conda environment:

conda create -n vsc python=3.8
conda activate vsc

3. Install dependencies:

pip install -r requirements.txt

Configuration

For Claude Desktop

Add this configuration to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "video-sum-mcp": {
      "command": "python",
      "args": ["/path/to/video-sum-mcp/main.py"],
      "cwd": "/path/to/video-sum-mcp",
      "env": {
        "CONDA_DEFAULT_ENV": "vsc"
      }
    }
  }
}

For Other MCP Clients

The server can be started directly:

python main.py

Usage

Basic Video Processing

# Example: Process a Bilibili video
result = process_video(
    url="https://www.bilibili.com/video/BV1234567890",
    output_format="markdown"
)

Supported URL Formats

• Douyin: https://v.douyin.com/... or full URLs
• Bilibili: https://www.bilibili.com/video/...
• Xiaohongshu: https://www.xiaohongshu.com/discovery/item/...
• Zhihu: https://www.zhihu.com/question/...

Context-Enhanced Processing

For platforms with anti-crawling measures, you can provide context:

result = process_video(
    url="https://...",
    context_text="Additional context information..."
)

Features in Detail

OCR Integration

• Automatic image text extraction from Xiaohongshu posts
• PaddleOCR for accurate Chinese character recognition
• Batch processing for multiple images

Knowledge Graph Generation

• Structured content analysis
• Intelligent relationship mapping
• Quality control and validation

Anti-Crawling Strategies

• Smart fallback mechanisms
• Context-based extraction
• User guidance for optimal results

Development

Project Structure

video-sum-mcp/
├── core/                 # Core functionality modules
│   ├── extractors/       # Platform-specific extractors
│   ├── processors/       # Content processing logic
│   ├── knowledge_graph/  # Knowledge graph generation
│   └── managers/         # Resource management
├── scripts/              # MCP server implementation
├── main.py              # Main entry point
├── requirements.txt     # Python dependencies
└── pyproject.toml       # Project configuration

Running Tests

python -m pytest

Dependencies

Key dependencies include:

• bilibili-api-python - Bilibili API integration
• yt-dlp - Video downloading capabilities
• PaddleOCR - OCR text recognition
• beautifulsoup4 - Web scraping
• requests - HTTP requests

See requirements.txt for complete list.

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Built using the Model Context Protocol
• OCR powered by PaddleOCR
• Platform integrations using various open-source APIs