WebMCP Lighthouse Score: What Is the Agentic Browsing Category?

What changed in Lighthouse 13.3

Google shipped Lighthouse 13.3 alongside Chrome 146 with a new audit category: Agentic Browsing. It sits alongside Performance, Accessibility, Best Practices, and SEO in the Lighthouse report panel in DevTools.

The category scores how ready a page is for AI agents — not just search crawlers. A page that registers WebMCP tools correctly can score 100. A page with no agent support scores 0.

What the Agentic Browsing audits check

The category contains four audits, each contributing to the 0–100 score:

Tool registration

Are WebMCP tools registered on navigator.modelContext? The audit checks for at least one registered tool with a valid name and description.

40 pts

Schema validity

Does each tool's inputSchema conform to JSON Schema draft-07? Tools with missing or malformed schemas fail this audit.

30 pts

ARIA consistency

Do the interactive elements that tools reference exist in the DOM with matching ARIA labels? Lighthouse cross-references tool descriptions against accessible names.

20 pts

Layout stability (CLS)

Does the page stay stable while tools execute? Unexpected layout shifts during agent interactions cause failures. Standard CLS rules apply.

10 pts

Why this score matters now

Google has confirmed that Agentic Browsing scores will eventually factor into search ranking signals, the same way Core Web Vitals did after their 2021 rollout. The rollout timeline follows the same pattern: announce → measure → weight.

Right now, the score is informational. Pages that get to 100 before weighting begins will have a head start — and the changes required are small if you use the right tooling.

How to check your current score

Open Chrome DevTools (Chrome 146+), run Lighthouse, and select the Agentic Browsing category. You'll see the four audits with pass/fail status and the contributing score for each.

If you don't have Chrome 146 yet, the Lighthouse CLI works too:

npx lighthouse https://yoursite.com --only-categories=agentic-browsing --output=html

How Aigentably gets you to 100

Aigentably handles the first two audits automatically. When you paste the Aigentably script tag, it:

Fetches your tool definitions from the Aigentably API
Registers them on navigator.modelContext with valid JSON Schema input schemas
Loads the WebMCP polyfill for browsers below Chrome 146

Tool registration and schema validity — 70 of the 100 points — are handled without any manual work. ARIA consistency and CLS depend on how your page is built, but Aigentably surfaces failures with specific guidance in the dashboard.