MCP.so
Sign In

Docs MCP Server

@arabold

RepositoryHomepage

About Docs MCP Server

Grounded Docs MCP Server: Open-Source Alternative to Context7, Nia, and Ref.Tools

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "docs-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@arabold/docs-mcp-server"
      ],
      "env": {
        "OPENAI_API_KEY": "sk-proj-..."
      },
      "autoApprove": []
    }
  }
}

Tools

No tools detected

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Docs MCP Server?

Docs MCP Server solves AI hallucinations and outdated knowledge by providing a personal, always-current documentation index for your AI coding assistant. It fetches official docs from websites, GitHub, npm, PyPI, and local files, allowing your AI to query the exact version you are using. It is an open-source alternative to Context7, Nia, and Ref.Tools.

How to use Docs MCP Server?

Install Node.js 22+ and use CLI commands: npx @arabold/docs-mcp-server@latest scrape <name> <url> to index docs, search <name> <query> to query, and fetch-url <url> to get a single page. For a long-running MCP endpoint, start the server with npx @arabold/docs-mcp-server@latest, open the Web UI at http://localhost:6280, and connect your MCP client (e.g., Claude, Cline) via SSE configuration. Optionally enable an embedding model (e.g., OpenAI, Ollama) to improve search quality.

Key features of Docs MCP Server

  • Up-to-date context fetched from official sources on demand
  • Version-specific queries target exact library versions
  • Reduces LLM hallucinations by grounding in real docs
  • Runs entirely locally; code never leaves your network
  • Works with any MCP-compatible client (Claude, Cline, etc.)
  • Supports websites, GitHub, npm, PyPI, local files, and zip archives

Use cases of Docs MCP Server

  • Querying the exact version of a library used in your project
  • Reducing hallucinations in AI coding assistant responses
  • Indexing private or local documentation alongside public sources
  • Searching across multiple documentation sources from one interface

FAQ from Docs MCP Server

What sources does Docs MCP Server support?

It fetches documentation from websites, GitHub repositories, npm, PyPI, local files, and zip archives. It processes HTML, Markdown, PDF, Office documents, and 90+ source code languages.

Does Docs MCP Server require an internet connection?

It fetches documentation on demand from official sources, so internet access is needed for remote sources. The server runs entirely on your machine; your code never leaves your network.

How do I configure embedding models?

Set environment variables like OPENAI_API_KEY to enable embeddings. Supported providers include OpenAI, Ollama, Gemini, and Azure. See the Embedding Models guide in the documentation.

Does Docs MCP Server support hash-routed SPA documentation sites?

Yes, use the --preserve-hashes flag (CLI) or preserveHashes: true (MCP tool) for sites that route with URLs like #/guide. This automatically upgrades the scraper to use Playwright.

What is the license for Docs MCP Server?

It is licensed under the MIT License. See the LICENSE file for details.

Frequently asked questions

What sources does Docs MCP Server support?

It fetches documentation from websites, GitHub repositories, npm, PyPI, local files, and zip archives. It processes HTML, Markdown, PDF, Office documents, and 90+ source code languages.

Does Docs MCP Server require an internet connection?

It fetches documentation on demand from official sources, so internet access is needed for remote sources. The server runs entirely on your machine; your code never leaves your network.

How do I configure embedding models?

Set environment variables like `OPENAI_API_KEY` to enable embeddings. Supported providers include OpenAI, Ollama, Gemini, and Azure. See the Embedding Models guide in the documentation.

Does Docs MCP Server support hash-routed SPA documentation sites?

Yes, use the `--preserve-hashes` flag (CLI) or `preserveHashes: true` (MCP tool) for sites that route with URLs like `#/guide`. This automatically upgrades the scraper to use Playwright.

What is the license for Docs MCP Server?

It is licensed under the MIT License. See the LICENSE file for details.

Comments

Basic information

Category

Memory & Knowledge

License

MIT

Transports

stdio

Author

arabold

Submitted by

Andre Rabold

More Memory & Knowledge MCP servers

Mcp Knowledge Graph

shaneholloman

MCP server enabling persistent memory for Claude through a local knowledge graph - fork focused on local development

871

mcp-local-rag

nkapila6

"primitive" RAG-like web search model context protocol (MCP) server that runs locally. ✨ no APIs ✨

128

Obsidian MCP Server

cyanheads

Read, write, search, and surgically edit Obsidian vault notes, tags, and frontmatter via MCP. STDIO or Streamable HTTP.

618

Notion MCP Server

awkoy

Notion MCP server for Claude, Cursor, ChatGPT & Claude Desktop. Connect AI agents to Notion via Model Context Protocol — pages, databases, blocks, comments, files.

159

Basic Memory

basicmachines-co

AI conversations that actually remember. Never re-explain your project to your AI again. Join our Discord: https://discord.gg/tyvKNccgqN

3.4K