Discoverability

Discoverability controls how easily AI agents and assistants can find, crawl, and understand your documentation. ReadMe crawls a sample of your published pages and scores your project against the Agent-Friendly Docs spec, so you can see at a glance where your docs stand and which switches to flip to improve the score.

Every setting on this page has a sensible default. Turn them on to expose more of your docs to agents, or leave them off if you'd rather keep a lower profile.

Getting Started

1. Open your project in ReadMe.
2. Go to AI → Discoverability.
3. Review your current grade and the individual settings under each group.
4. Toggle any setting on or off. Changes take effect on your published site.

Enterprise projects inherit these settings from their group by default. Change any of them on the project to override, or click Restore Parent Default on a setting to match the group again.

Content Discoverability

Help agents find and crawl your docs with llms.txt and link headers.

• LLMs.txt — Publishes an llms.txt file at the root of your docs (for example, your-project.readme.io/llms.txt) so AI assistants can read and understand your documentation.
• LLMs.txt: Split by Section — Splits llms.txt into per-section files (for example, /docs/llms.txt) with a small root index. Useful when llms.txt grows past size limits some AI agents enforce.
• LLMs.txt: Split by Category — Splits section files down to category-level files (for example, /docs/getting-started/llms.txt), with each section serving as a category index.
• Link Headers — Emits Link response headers (RFC 8288) pointing AI agents to your API catalog and other discovery endpoints.
• Agent Hint Banner — Injects a hidden in-page hint pointing AI agents at your llms.txt. In testing, this is the single fastest path from an agent's question to a useful answer.
• API Catalog — Publishes an RFC 9727 linkset at /.well-known/api-catalog listing every API definition on this project with links to each OpenAPI spec.
• Agent Skills Index — Publishes an Agent Skills Discovery index (v0.2.0) at /.well-known/agent-skills/index.json advertising skills specific to your documentation.

Content Signals

Control how AI companies may use your content — for training, search indexing, and prompt context.

Set each signal independently. Signals are advisory — they tell well-behaved crawlers what you allow.

Markdown Availability

Serve clean Markdown so agents ingest pages without parsing HTML.

• Markdown for Agents — When an agent sends Accept: text/markdown, ReadMe returns your docs as Markdown instead of HTML. HTML still ships by default for browsers.

Content Structure

Expose how your pages relate to each other so agents can follow the shape of your documentation.

• Show Sub Pages — Appends a list of child page links to each page's Markdown response so agents can crawl deeper.
• Show Sibling Pages — Appends a list of sibling page links so agents can move laterally between related pages.
• Show What's Next — Appends the author-curated "What's Next" links from page frontmatter to the Markdown response.

MCP

Let AI agents connect to your docs as tools — via a hosted MCP server and in-browser WebMCP. For the full MCP configuration surface (auth, custom tools, enabled routes), see MCP.

• MCP Server Card — Publishes an SEP-1649 Server Card at /.well-known/mcp/server-card.json advertising your project's MCP server.
• WebMCP Tools — Exposes four tools on every docs page via navigator.modelContext: search, navigate, read the current page, and try an API endpoint. In-browser AI agents can call them directly.

Authentication and Access

Advertise OAuth endpoints so agents can reach docs behind a login.

Set Discovery Type to control which OAuth discovery flavor is advertised on this site:

Leave this as None if your docs are public or you don't need agents to authenticate.

FAQ