docs-mcp-server
by MCP-Mirror
A MCP server for fetching and searching 3rd party package documentation. It scrapes, processes, indexes, and searches documentation for various software libraries and packages.
Last updated: N/A
What is docs-mcp-server?
The docs-mcp-server is a Model Context Protocol (MCP) server that scrapes, processes, indexes, and searches documentation for various software libraries and packages. It fetches content from specified URLs, splits it into meaningful chunks, generates vector embeddings using OpenAI, and stores the data in an SQLite database for efficient search.
How to use docs-mcp-server?
To use the server, you can either integrate it with Claude Desktop or run it using Docker. You can also interact with it using the provided CLI (docs-mcp). The server exposes MCP tools for starting scraping jobs, checking job status, listing jobs, cancelling jobs, searching documentation, listing indexed libraries, finding appropriate versions, and removing indexed documents.
Key features of docs-mcp-server
Scrapes and indexes documentation from various sources.
Uses semantic splitting and OpenAI embeddings for enhanced search.
Supports versioning of documentation.
Provides an asynchronous job queue for managing scraping processes.
Offers a CLI for local management and interaction.
Integrates with Claude Desktop for enhanced contextual understanding.
Use cases of docs-mcp-server
Providing documentation context to AI assistants like Claude.
Creating a searchable documentation repository for internal libraries.
Offering versioned documentation for different software releases.
Automating the process of keeping documentation up-to-date.
Enabling developers to quickly find relevant information within large documentation sets.
FAQ from docs-mcp-server
How do I start a scraping job?
How do I start a scraping job?
Use the scrape_docs MCP tool or the docs-mcp scrape CLI command, providing the library name and URL.
How do I search for documentation?
How do I search for documentation?
Use the search_docs MCP tool or the docs-mcp search CLI command, specifying the library and query.
How do I specify a version when scraping?
How do I specify a version when scraping?
Use the -v or --version option with the docs-mcp scrape command.
How do I search only unversioned documentation?
How do I search only unversioned documentation?
Use the --version "" option with the docs-mcp search command.
How do I integrate the server with Claude Desktop?
How do I integrate the server with Claude Desktop?
Add the server configuration to your claude_desktop_config.json file, specifying the command, arguments, and environment variables.