Agent readiness tools

WebMCP checker

Q: How do these checks count in the score?

webmcp is in the agent-conventions part: 30% of the score, and none of its checks holds a level. The methodology says what each state earns.

Enter a domain to read the WebMCP tools its homepage declares. The scan reads the HTML it is served and runs no JavaScript.

The WebMCP checker runs the full scan of 60 checks, not one request. The five exemplar scans of 20 October 2026 sent 29 to 89 requests each and took 3 to 8 seconds. A scan that reaches its 60 s deadline is reported as not measured.

A stored result: cloudflare.com, measured on 20 October 2026. Its report opens on the same panel yours will.

What is checked

The report opens on 1 check of rubric v0.9. Each is described here in the rubric's own words.

WebMCP browser tool declaration

WebMCP lets a page register tools for an in-browser agent. It is a W3C Community Group draft.

webmcp is a scored check. Agent conventions: 30% of the score. A convention never holds a level.

The scanner reads the homepage only: the WebMCP draft defines an in-page API and no manifest file. It counts registerTool() calls and forms that declare a tool with a toolname attribute. It is present when every tool is complete: a registerTool() call that names a name, a description, and an inputSchema, and a form that carries a toolname and a tooldescription. It is partial when a declared tool lacks one of them, and not used when the homepage has forms and declares no tool. A homepage with no form and no registerTool() call offers no task to declare, so the check is out of scope, as Lighthouse marks such a page. A complete declaration earns the check's points in the agent-conventions part, a partial one half of them, and none earns nothing.

Run the same measurement against your own site. Set the origin on the first line:

HOMEPAGE=https://example.com/
curl -sSL -A 'AgenticWebScore/0.3 (+https://agenticwebscore.com/bot)' -H 'Accept-Language: en' -H 'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,text/plain;q=0.8,*/*;q=0.7' "$HOMEPAGE" | grep -oE 'registerTool\(|<form[^>]*toolname[^>]*>' | sort | uniq -c

Read the webmcp check page for its sources and the change to make.

Questions about the WebMCP checker

Does the scan see a tool that a script registers?

The scanner reads the homepage only: the WebMCP draft defines an in-page API and no manifest file. It runs no JavaScript and reads no script file the page loads. So a registerTool() call in a script file is not counted, and one written in the homepage HTML is.

Is WebMCP a standard?

WebMCP is a W3C Community Group draft, not on the W3C Standards Track, and Chrome exposes it behind an origin trial. The draft defines the in-page API only, so there is no manifest file to publish.

How do these checks count in the score?

webmcp is in the agent-conventions part: 30% of the score, and none of its checks holds a level. The methodology says what each state earns.

Why does it run the full scan?

The form starts the same scan as /scan, so the result is a scan record with a fingerprint you can verify. The report opens on webmcp, and the full report follows. A stored record younger than 6 hours opens instead of a new scan.