Docy is a Model Context Protocol server that provides documentation access capabilities. This server enables LLMs to search and retrieve content from documentation websites by scraping them with crawl4ai. Built with FastMCP v2, Docy exposes tools and prompts to discover documentation sources, fetch full pages as markdown, and extract links for navigation. It supports integration with Claude, Code, and other MCP-enabled tools, allowing AI models to access up-to-date documentation directly during tasks. Sources can be configured via DOCY_DOCUMENTATION_URLS or a .docy.urls file for hot reloading, with a disk-based cache to speed responses and reduce external requests. The server pre-fetches configured sites at startup and caches content with a configurable TTL, while the .docy.urls approach re-reads sources on each request to support dynamic changes. Docy can be run via uv, pip, or Docker, and supports SSE or stdio transports. The server is self-hosted and configurable, enabling teams to maintain secure, perimeters-contained documentation access.