WebMCP registry

WebMCP registry: sites with WebMCP and MCP discovery signals

PurposeBot crawls the web for sites that expose tools to AI agents: WebMCP page tools, MCP manifests and registry feeds, and OpenAPI documents. Each domain below has its own page showing exactly what the crawler found, so agents and developers can judge how much to trust it. This is the human-readable view of the public registry feed (JSON).

WebMCP and MCP solve related problems. The W3C WebMCP draft lets a web page register tools for agents in the browser (document.modelContext). MCP connects agents to tool servers. PurposeBot's /.well-known/webmcp document is a PurposeBot discovery endpoint, not the official WebMCP page API. Listing here is an automated observation, not an endorsement or a compliance claim.

Domains: 371
URLs checked: 2,000
Machine documents / MCP registry: 24
Feed generated: 2026-10-09

Search the agent registry Live registry explorer How PurposeBot trust scores work

Machine documents and MCP registry listings (24)

The crawler retrieved a machine-readable discovery document, MCP manifest, registry feed, or published API endpoints for this domain, or the MCP registry lists this domain's site.

Sites whose pages reference WebMCP (14)

Page HTML on this domain references WebMCP (for example "webmcp" or a /.well-known/webmcp link), but no machine-readable WebMCP document was retrieved.

Low-evidence entries (333)

Only indirect probe results were observed (access-protected or inconclusive responses). This entry is listed for completeness and is not indexed.