Grade any page the way an AI agent reads it
Open a page, open the side panel, get a score. Yours, a competitor’s, a prospect’s — it works on anything you can load in a tab.
Agent Readiness, 0–100
Search engines index pages. Agents have to use them — find the page, parse it, work out what actions are available, and get through whatever stands in the way. Four pillars, 25 points each.
Discovery
Can an agent find the page at all — sitemaps, crawler permissions, canonical signals, internal links.
Machine readability
Once it has the page, can it parse it — structured data, answer-shaped content, headings that mean something, machine-readable feeds.
Capabilities
Can it tell what you can do here — the actions, the parameters, the prices, the availability. This is the pillar most sites score worst on.
Friction
And what stops it — walls, scripts that hide the content, interstitials, anything that answers a bot differently from a browser.
A grade, and the reason for it
Every failing check comes with what it is, why it matters and what to change. Print it or hand it to your coding agent.
The diagnosis is free and always will be. Every check, every score, every explanation, plus the generic fix for the SEO and GEO checks. What is not free is the site-specific work: directives written against your actual pages, and someone reading the result.
The extension grades one page. I read the whole picture.
The extension tells you what a page is missing. An audit tells you which questions your buyers ask, who gets named instead of you across six engines, and which fixes are worth doing first.