Docs Fetch MCP Server
by wolfyy970
A Model Context Protocol (MCP) server for fetching web content with recursive exploration capabilities. This server enables LLMs to autonomously explore web pages and documentation to learn about specific topics.
Last updated: N/A
What is Docs Fetch MCP Server?
The Docs Fetch MCP Server is a tool that allows LLMs to retrieve and explore web content. It fetches clean, readable content from web pages and recursively explores linked pages within the same domain to gather comprehensive information.
How to use Docs Fetch MCP Server?
The server exposes a single MCP tool called fetch_doc_content. You provide a URL and an optional depth parameter (1-5) to specify how many levels of linked pages to explore. The server returns a JSON object containing the explored pages, their content, and links.
Key features of Docs Fetch MCP Server
Content Extraction
Link Analysis
Recursive Exploration
Parallel Processing
Robust Error Handling
Dual-Strategy Approach
Timeout Prevention
Partial Results
Use cases of Docs Fetch MCP Server
Enabling LLMs to learn about specific topics by exploring documentation
Gathering comprehensive information from websites for LLM training
Automating web content research for LLMs
Providing LLMs with context from web pages for improved reasoning
FAQ from Docs Fetch MCP Server
What is the default exploration depth?
What is the default exploration depth?
The default exploration depth is 1.
What is the maximum exploration depth?
What is the maximum exploration depth?
The maximum exploration depth is 5.
What happens if a page fails to load?
What happens if a page fails to load?
The server implements robust error handling and will attempt to return partial results, providing available content even if some pages fail to load completely.
What dependencies are required?
What dependencies are required?
The server depends on @modelcontextprotocol/sdk, puppeteer, and axios.
How does the server handle complex web pages?
How does the server handle complex web pages?
The server uses a dual-strategy approach, first attempting to fetch content with fast axios requests. If that fails, it falls back to puppeteer for more complex pages.