Docfork Mcp

docfork/docfork-mcp

Gives your AI coding agent access to current documentation for 9000+ libraries without polluting context with stale snippets. Exposes two tools: search_docs returns ranked sections with relevance scores, fetch_doc pulls full markdown content from a URL. Solves the problem of agents hallucinating deprecated APIs or using outdated signatures because their training data is old. Run npx dgrep setup to install, then your agent can search Next.js App Router docs, Hono middleware patterns, or any public library on the fly. You can also index private repos so internal SDKs and runbooks become searchable the same way. Free tier includes 1,000 requests per month per org.

GitHub stars: 487
tools: 2
auth: Secret needed
transport: HTTP
last updated: Sep 12, 2025

View on GitHub

Tools

2 tools

Public tool metadata for what this MCP can expose to an agent.

docfork_search_docsSearch documentation across GitHub repositories or the web. For targeted searches INSIDE a specific library's documentation, use the docforkIdentifier parameter (author/repo format). Extract from GitHub URLs (e.g., github.com/facebook/react → 'facebook/react') and include in A...3 params

Search documentation across GitHub repositories or the web. For targeted searches INSIDE a specific library's documentation, use the docforkIdentifier parameter (author/repo format). Extract from GitHub URLs (e.g., github.com/facebook/react → 'facebook/react') and include in A...

Parameters* required

docforkIdentifierstring

CRITICAL for targeted library searches: Library identifier in author/repo format (e.g., 'facebook/react', 'vercel/next.js'). Use this to search INSIDE a specific library's documentation for focused...

querystring

Search query. Include language/framework names.

tokensstring

Token budget: 'dynamic' or number (100-10000)

docfork_read_urlFetches and returns the full content of a documentation page as markdown. This is ESSENTIAL for getting complete, detailed information after searching. Always use this to read URLs from docfork_search_docs results to get the actual documentation content.1 params

Fetches and returns the full content of a documentation page as markdown. This is ESSENTIAL for getting complete, detailed information after searching. Always use this to read URLs from docfork_search_docs results to get the actual documentation content.

Parameters* required

urlstring

Full URL to read. Use exact URLs from docfork_search_docs results.

Readme

Docfork

[!CAUTION] Docfork has shut down as of June 14, 2026. The MCP server, CLI, and all hosted endpoints are offline and will not return. API keys no longer work, and npx dgrep setup will fail. Thank you to everyone who used the service.

Up-to-date docs for AI coding agents.

AI agents hallucinate APIs, bloat context with stale docs, and write code against outdated signatures. Docfork serves up-to-date documentation directly in Cursor, Claude Code, and Windsurf.

Without Docfork

  app.use('/api/*', jwt({ secret: ... }))
-                       ^^^ removed in Hono v4

With Docfork

  app.use('/api/*', bearerAuth({ verifyToken: ... }))
+                       ^^^ current API, Hono v4.2

Get Started

npx dgrep setup

View the full README on GitHub

Details

Repository: github.com/docfork/docfork-mcp
Transport: HTTP
Auth: Required
Registry: MCP registry, active