Ragdocs MCP Server

Vector-based documentation search and indexing for AI-driven knowledge retrieval.

Data last scanned 2 days ago · Reviewed 2 days ago

28 2 dangerous tools7 tools★ 265TypeScriptMIT

Overview

Ragdocs MCP Server allows AI assistants to perform semantic search across documentation sources. By indexing content through a vector-based approach, it enables models to retrieve relevant excerpts and contextual information to augment their responses.

Our verdict

This server carries a high-risk rating with a security score of 28/100. The repository has not seen a commit in over a year, and it lacks an official editorial review, making it unsuitable for environments requiring long-term maintenance or verified security standards.

Supports semantic search across multiple documentation sources
Provides seven direct dependencies for a reduced supply-chain surface
Includes tools to extract, queue, and index URLs

Last repository commit occurred over 369 days ago
Lacks an official vendor maintainer for support
Authentication methods remain unreviewed

Setup

Claude Desktop

{
  "mcpServers": {
    "mcp-ragdocs": {
      "args": [
        "-y",
        "@hannesrudolph/mcp-ragdocs"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
search_documentation	Search through stored documentation using natural language queries.	low
list_sources	List all documentation sources currently stored in the system.	low
extract_urls	Extract and analyze all URLs from a given web page.	medium
remove_documentation	Remove specific documentation sources from the system by their URLs.	high
list_queue	List all URLs currently waiting in the documentation processing queue.	low
run_queue	Process and index all URLs currently in the documentation queue.	medium
clear_queue	Remove all pending URLs from the documentation processing queue.	high

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›What prerequisites are required to run the RAG Documentation MCP server?

You must provide an OpenAI API key for generating embeddings and a Qdrant vector database instance with a corresponding URL and API key.

›How do I limit the number of search results returned?

The search_documentation tool accepts an optional limit parameter, which accepts a number between 1 and 20, with a default of 5 results.

›Does the extract_urls tool support local files?

No, the extract_urls tool requires a complete URL starting with a protocol like https:// and the target page must be publicly accessible.

›How are documentation sources removed from the system?

You can use the remove_documentation tool by providing an array of URLs that exactly match the URLs used when the documentation was originally added.

›What is the difference between the documentation queue and the indexed database?

The queue holds pending URLs waiting for indexing, while the indexed database stores documentation ready for semantic search. You can process the queue using the run_queue tool.