mcp-server-fetch

Fetch MCP Server is designed to enable Language Model Models (LLMs) to retrieve and process content from web pages, even those requiring JavaScript rendering or employing techniques that prevent simple scraping. It uses a combination of browser automation, OCR, and various extraction methods to ensure comprehensive content retrieval. The server employs a sophisticated scoring system to select the best content, ensuring high-quality and reliable results. Debug logging is available to track scoring decisions, making it a robust tool for content extraction.

Features

• Browser automation with undetected-chromedriver for dynamic content rendering.
• OCR using pytesseract with layout detection for text extraction from images.
• HTML extraction using requests and BeautifulSoup for static content.
• Document parsing capabilities for formats like PDF, DOCX, and PPTX.
• Sophisticated scoring system to ensure high-quality content selection.

MCP Tools

• {'fetch': 'Fetches a URL from the internet using browser automation and multi-method extraction, including OCR.'}

Usage with Different Platforms

docker_installation

To install and run mcp-server-fetch using Docker, follow these steps:

1. **Build the Docker image:**
   bash
   docker build -t mcp-server-fetch .
   

2. **Run the Docker container:**
   bash
   docker run --rm -i mcp-server-fetch
   

claude_configuration

Add to your Claude settings:


{
  "mcpServers": {
    "fetch": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "mcp-server-fetch"
      ],
      "disabled": false,
      "alwaysAllow": []
    }
  }
}