maartensmeets_mcp_server_fetch
by MaartenSmeetsFetch MCP Server: Web Content Retrieval for AI Models
The Fetch MCP Server is a specialized tool designed to retrieve and process web content for AI models. It leverages browser automation, OCR, and advanced extraction methods to fetch information from URLs, even for pages requiring JavaScript rendering or anti-scraping techniques.
Overview
The Fetch MCP Server is a Model Context Protocol (MCP) implementation that enables AI models to access and process web content efficiently. It supports multiple extraction methods, including browser automation, OCR, and document parsing, ensuring high-quality results for complex web pages.
Key Features
- Multi-Method Extraction: Combines browser automation, OCR, HTML extraction, and document parsing.
- Sophisticated Scoring System: Evaluates content quality based on length, structure, and readability.
- Advanced Automation: Handles cookie consent banners, captures full-page screenshots, and performs OCR with layout detection.
- Customizable User-Agent: Allows customization of the user-agent string for requests.
- Docker Support: Easy deployment using Docker.
Available Tools
fetch Tool
- Description: Fetches web content from a specified URL using browser automation and multi-method extraction.
- Arguments:
url(string, required): The URL to fetch.raw(boolean, optional): Retrieves raw HTML content without simplification (default:false).
Installation
Using Docker
-
Build the Docker Image:
shell docker build -t mcp-server-fetch . -
Run the Docker Container:
shell docker run --rm -i mcp-server-fetch
Configuration
Configure for Claude App
Add the following to your Claude settings:
{
"mcpServers": {
"fetch": {
"command": "docker",
"args": [
"run",
"--rm",
"-i",
"mcp-server-fetch"
],
"disabled": false,
"alwaysAllow": []
}
}
}
Customize User-Agent
Modify the user-agent string by adding the --user-agent=YourUserAgent argument to the args list in the configuration.
Contributing
We welcome contributions to enhance the Fetch MCP Server. Whether you want to add new tools, improve existing functionality, or refine documentation, your input is valuable.
For examples of other MCP servers, visit: https://github.com/modelcontextprotocol/servers
License
The Fetch MCP Server is licensed under the MIT License. For more details, see the LICENSE file.
Invite Gmail Contacts
