Docs MCP Server

An MCP (Model Context Protocol) server that exposes Markdown documentation, API reference (C# XML-doc / TypeDoc), and JSON Schema / OpenAPI specs to AI agents. From local folders or GitHub repositories.

A single server instance can host multiple libraries/frameworks side-by-side, each with its own sources. The AI picks which library to query via a library parameter, discoverable through the list_libraries tool.

Why use this instead of direct file access?

| | Direct file access | Docs MCP | |---|---|---| | Security | Agent can read/write anywhere on the filesystem | Sandboxed per source with traversal protection | | Discovery | Agent scans directories and reads files one-by-one | Index tools give instant overviews of every source | | Search | Agent greps files manually, burning context | Dedicated search tools with regex or glob support | | Large files | Entire file loaded into context | TOC + chapter extraction reads only needed sections | | Multi-library | Agent must know every path/repo | One server, many libraries, self-describing | | Source | Local files only | Local directories or GitHub URLs — no cloning required |

Quick start

npm install
npm run build

# Single library
docs-mcpserver ./docs --name "My Docs"

# Multi-library
docs-mcpserver --config dev-docs.json

See sample-config.json for a complete multi-library example.

Documentation

• Getting started — install, CLI, Claude Code / Claude Desktop integration
• Configuration — config file format, libraries, sources
• Tools — the tool groups the agent sees (docs / api / schema)
• Hosting — HTTP mode and IIS (httpPlatformHandler) setup

License

AGPL