Skip to main content

Universal MCP server for documentation with llms.txt support

Project description

docr-mcp

docr-mcp

A framework for building MCP servers that give LLMs access to any documentation.

Give LLMs the ability to search and read documentation from any source - public or private, official or internal. Stop getting outdated answers. Start getting accurate information directly from current docs.

Tests Python 3.10+ License: MIT

Why docr-mcp?

The Problem: LLMs give you outdated answers based on their training data. They don't know about the latest API changes, new features, or the specific libraries and tools you use daily.

The Solution: docr-mcp provides a proven framework to connect any documentation to LLMs through MCP servers. Give your AI assistant real-time access to current documentation - from popular libraries to your internal tools.

Key Features:

  • Universal framework for any documentation site (public or private)
  • Smart BM25 search with code-aware tokenization and relevance ranking
  • Full customization - control parsing, indexing, search, and tool descriptions
  • Production ready - 88 tests (unit + E2E), secure by default, proper resource management
  • Easy to extend - YAML config + Python implementation to add any library

Supported Documentation

Library Status Tests Install Command (Claude Code)
Strands Agents Test Strands claude mcp add docr-mcp-strands -- uvx docr-mcp --library strands
Vercel Test Vercel claude mcp add docr-mcp-vercel -- uvx docr-mcp --library vercel
Twilio Test Twilio claude mcp add docr-mcp-twilio -- uvx docr-mcp --library twilio
OpenAI Test OpenAI claude mcp add docr-mcp-openai -- uvx docr-mcp --library openai
Stripe Test Stripe claude mcp add docr-mcp-stripe -- uvx docr-mcp --library stripe
Anthropic Test Anthropic claude mcp add docr-mcp-anthropic -- uvx docr-mcp --library anthropic

Want to add a library? See CONTRIBUTING.md for guidelines.

Usage

Public Documentation

After installation, ask your AI assistant to search documentation:

Search Strands docs for "agent state"
What is agent-loop in Strands?
Show me how to use model providers in Strands

Authenticated Documentation (Private/Internal Docs)

Need to access private or internal documentation behind authentication? docr-mcp supports cookie-based authentication for SSO-protected sites.

Simply create a YAML config file with your cookies and domain - no code changes required.

📖 See the Authenticated Docs Guide for step-by-step instructions.

How It Works

graph LR
    A[Index Source] --> B[Build BM25 Index]
    B --> C[Query]
    C --> D[Search Index]
    D --> E[Fetch Live Docs]
    E --> F[LLM Response]
  1. Startup: Fetch index source (llms.txt, sitemap) and build searchable BM25 index
  2. Query: Search pre-built index, then fetch live documentation from URLs
  3. Search: BM25 ranking with code-aware tokenization and field weighting

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

docr_mcp-0.3.0.tar.gz (22.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

docr_mcp-0.3.0-py3-none-any.whl (4.0 kB view details)

Uploaded Python 3

File details

Details for the file docr_mcp-0.3.0.tar.gz.

File metadata

  • Download URL: docr_mcp-0.3.0.tar.gz
  • Upload date:
  • Size: 22.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for docr_mcp-0.3.0.tar.gz
Algorithm Hash digest
SHA256 f13c2922f4553683e0801a27a649753598b1129ecdd8da35a39f54b9dc724a79
MD5 6ecbbcb968357876643b6b336a27129c
BLAKE2b-256 ccbd691c6708400890a24065ad84d95ac1c42e729d7f57cfec9df36d297dbc91

See more details on using hashes here.

Provenance

The following attestation bundles were made for docr_mcp-0.3.0.tar.gz:

Publisher: publish.yml on JacobHuang91/docr-mcp

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file docr_mcp-0.3.0-py3-none-any.whl.

File metadata

  • Download URL: docr_mcp-0.3.0-py3-none-any.whl
  • Upload date:
  • Size: 4.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for docr_mcp-0.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 318e8cd326c92c62a07d33adf0ef9b23aefc3d4ca39a288ce9dea2bb9fd8758b
MD5 89e5b05855a7c08efaae6f25dc8a7231
BLAKE2b-256 07a5d957d2e7805912cd43308d04bcca004ff9051d794c0ca3ef7e95fefc0db7

See more details on using hashes here.

Provenance

The following attestation bundles were made for docr_mcp-0.3.0-py3-none-any.whl:

Publisher: publish.yml on JacobHuang91/docr-mcp

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page