Learn

Does the homepage return useful discovery metadata in HTTP `Link` response headers, letting agents find resources without parsing HTML?

Is there a valid `/robots.txt`, does it point to at least one sitemap, and does it address AI bots (either by `User-Agent` or by `Content-Signal`)?

Does the site publish a valid sitemap (directly at `/sitemap.xml`, via `robots.txt`, or as a sitemap index), and is it well-formed?

Does the site publish a well-formed `/llms.txt` that gives an LLM a compact, structured reading list?

When an agent asks for `Accept: text/markdown`, does the server return a clean markdown version of the page instead of HTML?

Does the homepage embed at least one well-formed JSON-LD block describing the site, org, or key content?

Does `robots.txt` have at least one stanza explicitly addressing a known AI bot (allow or disallow), showing the owner has made a considered decision?

Does `robots.txt` include a `Content-Signal` directive declaring AI training / input / search preferences?

Does the site publish a `/.well-known/http-message-signatures-directory` allowing bots to cryptographically authenticate requests to it (or, if this site runs its own bots, advertise its public keys for receivers to verify)?

Does the site expose an Agent-to-Agent (A2A) Agent Card at `/.well-known/agent-card.json`?

Does the site publish an Agent Skills index at `/.well-known/agent-skills/index.json`, listing skill documents agents can download and use?

Does the site publish an API Catalog at `/.well-known/api-catalog`, listing all its public APIs in one discoverable place?

Does the site expose an MCP Server Card at `/.well-known/mcp/server-card.json` so MCP-capable agents can discover, describe, and connect to its MCP server without prior knowledge?

Does the site expose OAuth authorization-server metadata at `/.well-known/oauth-authorization-server`, so agents can bootstrap authorization flows without hardcoded endpoints?

Does the site expose `/.well-known/oauth-protected-resource` metadata so agents can discover which authorization server protects a given resource?

Does the site expose a WebMCP endpoint — an in-page mechanism letting agents in a browser invoke MCP tools provided by the page itself?

Does the site expose ACP metadata at `/.well-known/agentic-commerce.json`, enabling agents to complete purchases on behalf of humans without parsing checkout HTML?

Does the site expose UCP discovery metadata, letting agents find product catalogs and checkout flows programmatically?

Does the site respond to an agent request with HTTP 402 and a machine-readable payment payload, signalling support for the x402 agentic commerce standard?

Does the site publish a `/humans.txt` crediting the people behind it?

Is an OpenAPI (or AsyncAPI) specification discoverable at a well-known location or via the API Catalog?

Does the homepage advertise an RSS or Atom feed via `<link rel="alternate" type="application/rss+xml|atom+xml">` or HTTP Link header?

Does the site publish `/.well-known/security.txt` with at least a `Contact:` and `Expires:` field?