Lighthouse Now Audits Your Site for AI Agents. Here's How to Pass the Test.
Lighthouse AI Agent AuditLighthouse Now Audits Your Site for AI Agents. Here's How to Pass the Test.
Chrome just shipped an experimental Lighthouse category called Agentic Browsing. It scores how well an AI agent — ChatGPT's Operator, Claude with computer use, Perplexity's assistant, a custom GPT — can actually read your page, find the right button, and complete a task on your site. For two years, "AI search" has been a fuzzy marketing topic measured by screenshots and vibes. Now Google has put it inside the same tool your dev team already runs on every deploy. If your buyers are starting to send agents instead of clicks, this audit is the first real scorecard for whether your site is ready to receive them. Here's what Lighthouse checks, what it ignores, and the five fixes that move the needle this quarter.







