Google Lighthouse Now Scores Your Site for AI Agents
Google released Lighthouse Agentic Browsing — a new experimental audit that checks whether AI agents can navigate, understand, and take action on your website. Four checks. Pass or fail. Here's everything you need to know.
TL;DR
- Google Lighthouse added an "Agentic Browsing" category (experimental)
- It checks 4 things: WebMCP tools, accessibility tree, layout stability (CLS), and llms.txt
- Score is pass/fail fractions, not 0-100
- Causabi now runs all 4 checks in your AI Readiness audit automatically
What is Lighthouse Agentic Browsing?
You know Lighthouse — Google's tool that scores your site on Performance, Accessibility, Best Practices, and SEO. Now there's a fifth category: Agentic Browsing.
The premise is simple: AI agents (like ChatGPT, Gemini, Perplexity, and browser-integrated AI assistants) interact with websites differently than humans do. They rely on the accessibility tree instead of visual layout, they try to find machine-readable summaries before crawling, and they need stable layouts to reliably click buttons and fill forms.
Lighthouse Agentic Browsing measures how ready your site is for this new type of visitor.
The 4 Agentic Browsing Audits
1. WebMCP Tool Registration
NEWWebMCP is a Chrome-proposed standard that lets your site declare its interactive capabilities to AI agents. Instead of an AI agent guessing how to book an appointment or add something to a cart, your forms explicitly announce what they do.
Lighthouse audits both the declarative version (HTML attributes on forms) and the imperative version (navigator.modelContext.registerTool() in JavaScript). The audit is informational — it lists registered tools and flags if key forms are missing declarations.
2. Agent-Focused Accessibility
PASS/FAILAI agents navigate your site using the accessibility tree — the same structure screen readers use. If your interactive elements lack labels, the agent simply cannot find or interact with them.
Lighthouse checks three things: names and labels on interactive elements, correct roles and parent-child relationships in the DOM, and visibility of content during agent interaction. The documentation puts it bluntly: "Missing labels can block users with visual disabilities and agents from completing a task."
3. Layout Stability (CLS)
PASS/FAILCumulative Layout Shift (CLS) was always a Core Web Vital for user experience. For AI agents, it's even more critical: if a button moves between when the agent identifies it and when it clicks, the interaction fails.
Target: CLS < 0.1. Common causes of failures: unsized images, late-loaded ads or cookie banners, and injected content above existing elements.
4. llms.txt Discoverability
PASS/FAILGoogle officially describes llms.txt as "an emerging convention used to provide a machine-readable summary of a website's content, specifically designed for LLMs and AI agents." It's now a Lighthouse Agentic audit.
The audit checks for the file at your domain root. Server errors are flagged as issues; 404 (file not found) is marked "Not Applicable" since implementation is still optional — though Google including it in Lighthouse is a strong signal this will become standard practice.
How the Score Works (It's Not 0-100)
Unlike every other Lighthouse category, Agentic Browsing does not produce a weighted 0-100 score. Instead, you get a fractional ratio: how many of the readiness checks pass.
Google is explicit about why: agentic web standards are still being developed. The category is in data-collection mode — providing useful signals without locking in a scoring methodology that might change as WebMCP and related standards mature.
This also means the score can fluctuate: dynamic WebMCP tool registration via JavaScript may be missed during Lighthouse snapshots, and accessibility tree changes from DOM modifications affect results run-to-run.
What to Do Right Now
Add llms.txt to your domain root
Create a Markdown file at /llms.txt with your business description, key pages, and contact information. Follow the llms-txt.org specification. This is the fastest win — takes 10 minutes.
Add WebMCP declarations to your forms
Add toolname and tooldescription attributes to key <form> elements — your contact form, booking form, search, add-to-cart. This is declarative WebMCP and Lighthouse will detect it immediately.
Audit your image alt text
Every <img> needs a descriptive alt attribute. Not empty alt="" (which hides images from agents), but actual descriptions. AI agents use alt text to understand visual content.
Fix semantic HTML landmarks
Ensure your pages use <header>, <main>, <nav>, and <footer>. The accessibility tree agents rely on is built from these semantic elements.
Measure and fix CLS
Run PageSpeed Insights on your homepage. If CLS > 0.1, the most common fixes are: add width/height to images, avoid late-injected content above the fold, preload critical resources.
Causabi Now Checks All 4 Lighthouse Agentic Audits
We've integrated all four Lighthouse Agentic Browsing checks into the Causabi AI Readiness audit. When you run an audit, you now get an Agentic Readiness section alongside your 0-100 score:
- +llms.txt presence and quality check
- +Image alt text audit across all crawled pages
- +Semantic HTML landmarks (<header>, <main>, <nav>, <footer>)
- +ARIA label detection on interactive elements
- +WebMCP declarative tool detection on forms
- +CLS measurement via PageSpeed Insights API
Every failing check comes with a specific fix — including the exact HTML attributes to add for WebMCP tool registration.
Free AI Readiness Check
See your GEO score before reading
Instant score · No email required · Checks ChatGPT, Gemini, Grok signals
The Bigger Picture
The release of Lighthouse Agentic Browsing is significant for one reason beyond the audits themselves: it signals that Google is treating AI agent interaction as a first-class web standard, not an experiment.
When Google adds something to Lighthouse, it enters the vocabulary of every developer, every SEO tool, every CMS plugin. Just as Performance scores drove adoption of lazy loading and Core Web Vitals drove image optimization, Agentic Browsing scores will drive adoption of WebMCP, llms.txt, and accessible markup as baseline requirements.
The window to get ahead of this is now — before it becomes another checkbox everyone scrambles to check retroactively.
Related Articles
FAQ
What is Lighthouse Agentic Browsing?
Lighthouse Agentic Browsing is an experimental audit category added to Google Lighthouse that evaluates how well your website is prepared for AI agents. Unlike the traditional 0-100 Lighthouse score, it uses a fractional pass/fail system across four areas: WebMCP tool registration, agent-focused accessibility, layout stability (CLS), and machine-readable discoverability via llms.txt.
How is the Lighthouse Agentic score calculated?
Unlike traditional Lighthouse categories, the Agentic Browsing category does not produce a 0-100 weighted score. Instead, it shows a fractional ratio (e.g., 3/5 checks passing), pass/fail status per audit, and informational counts. Google explicitly states this is because agentic web standards are still under development.
What is WebMCP and how do I add it to my site?
WebMCP is a Chrome-proposed standard that lets websites declare their interactive capabilities to AI agents. The declarative version requires adding toolname and tooldescription HTML attributes to your <form> elements — for example: <form toolname='book_appointment' tooldescription='Book a consultation with our team'>. The imperative version uses navigator.modelContext.registerTool() in JavaScript.
Does llms.txt affect the Lighthouse Agentic score?
Yes. llms.txt is explicitly listed as a Lighthouse Agentic Browsing audit under the Discoverability category. The audit checks whether a machine-readable summary file exists at your domain root. Server errors are flagged; 404 responses are marked 'Not Applicable' since the file is optional.
How does Causabi help with Lighthouse Agentic Browsing?
Causabi now includes all four Lighthouse Agentic Browsing checks in its AI Readiness audit: llms.txt generation, image alt-text audit, semantic HTML landmark detection, ARIA label verification, WebMCP tool declaration checking, and CLS measurement via PageSpeed Insights. Sites audited with Causabi receive actionable fixes for each failing check.