Skip to main content
You can read Astronomer’s documentation the way you always have, by browsing the site. If you’re working with an AI agent, you have three additional ways to bring this documentation into its context: a hosted MCP server, raw Markdown pages, and llms.txt index files.

​
MCP server

Astronomer hosts a Model Context Protocol (MCP) server at https://www.astronomer.io/docs/mcp. Connect an MCP-compatible client to it and the agent can search this documentation and pull specific pages into its context as it works, instead of relying on what it already knows.
Opening https://www.astronomer.io/docs/mcp in a browser returns a large JSON file. That’s expected — it’s an API endpoint for MCP clients, not a page meant to be viewed directly.
The server exposes three tools:
  • search_astronomer — full-text search across the documentation and API references.
  • query_docs_filesystem_astronomer — read-only, shell-like queries against a virtual filesystem of every documentation page and OpenAPI spec.
  • submit_feedback — report a page that’s incorrect, outdated, confusing, or incomplete.

​
Connect a client

​
Markdown pages

Every documentation page is also available as raw Markdown. Append .md to any page’s URL to fetch its plain-text source instead of the rendered HTML, for example https://www.astronomer.io/docs/astro/overview.md. On the page itself, use the contextual menu in the top right to copy the page as Markdown or open it directly in ChatGPT or Claude.

​
llms.txt

The root llms.txt file is an index of pointers to each product’s documentation, in the format proposed by the llms.txt standard. Each product has its own nested llms.txt, and versioned products like Astro Private Cloud and the Astro CLI nest one level deeper, with a file per version:
These files are generated to cover every supported and archived version of every product, including versions that aren’t the default on the site.