Multi Fetch MCP Server
by alexyangjie
This server extends the Fetch MCP Server by Anthropic, providing web content fetching capabilities. It allows LLMs to retrieve and process web page content, converting HTML to markdown and supporting chunked reading via start index.
Last updated: N/A
What is Multi Fetch MCP Server?
The Multi Fetch MCP Server is a tool that enables Large Language Models (LLMs) to fetch and process content from web pages. It converts HTML content to markdown for easier consumption by LLMs and supports fetching multiple URLs concurrently.
How to use Multi Fetch MCP Server?
The server is configured within an LLM environment like Claude, specifying the command to execute the server (e.g., using uvx). The LLM can then use the fetch or fetch_multi tools, providing URLs and optional parameters like max_length, start_index, and raw to control the content extraction.
Key features of Multi Fetch MCP Server
Fetches web content from URLs.
Converts HTML to markdown.
Supports fetching multiple URLs concurrently.
Allows specifying a start index for chunked content retrieval.
Offers options for raw content retrieval without markdown conversion.
Customizable user-agent and proxy settings.
Can be configured to respect or ignore robots.txt.
Use cases of Multi Fetch MCP Server
Enabling LLMs to access and process real-time information from the web.
Providing LLMs with context from web pages for improved responses.
Building LLM-powered web scraping and data extraction tools.
Creating LLM agents that can browse the web and gather information autonomously.
FAQ from Multi Fetch MCP Server
How do I configure the server to use a proxy?
How do I configure the server to use a proxy?
Use the --proxy-url argument followed by the proxy URL (e.g., --proxy-url http://192.168.1.1:8080 for HTTP or --proxy-url socks5://192.168.1.1:8080 for SOCKS5).
How can I customize the user-agent?
How can I customize the user-agent?
Use the --user-agent argument followed by your desired user-agent string (e.g., --user-agent=MyCustomAgent).
How do I ignore robots.txt?
How do I ignore robots.txt?
Add the --ignore-robots-txt argument to the server's configuration.
What is the default maximum length of the fetched content?
What is the default maximum length of the fetched content?
The default maximum length is 50000 characters.
How can I debug the server?
How can I debug the server?
Use the MCP inspector with the command npx @modelcontextprotocol/inspector uvx mcp-server-multi-fetch (or the appropriate command for your installation method).