Your docs as a tool for every AI agent
Every published site will get its own Model Context Protocol server, so agents in editors and chat apps can search and read your docs directly.
The per-site MCP server is on the roadmap and isn’t available yet. Today, agents can read any page as Markdown, fetch llms.txt and llms-full.txt, and use the REST API with a personal token.
What we’re designing for
Search your docs
A tool that runs the same search readers use, returning the best-matching pages.
Read any page
A tool that returns a page as Markdown, straight from your published content.
No setup
Generated when you publish, with the same visibility as the site itself.
Use your docs from agents today
Point any agent that can fetch URLs at your site’s Markdown files.
# Give an agent the whole site as context
curl https://docs.example.com/llms-full.txt
# Or fetch exactly the page it needs
curl https://docs.example.com/api/authentication.mdFrequently asked questions
What is MCP?
The Model Context Protocol is an open standard that lets AI assistants call tools and read data from outside sources in a consistent way.
Is the MCP server available?
Not yet. It’s on Docsheaf’s roadmap, and we don’t have a date to share.
Which clients will it work with?
It’s designed to work with any MCP-compatible client.
How can agents use my docs now?
Through Markdown versions of every page, llms.txt and llms-full.txt, or the REST API with a personal token.
Give agents docs worth reading
Publish accurate, reviewed docs today — the MCP server will serve the same content.