Can an AI read your site?
We scanned 300 marketing and SEO agency homepages the way a language model reads them. 247 returned a score. 52 would not let us read them at all — a robots.txt rule, an HTTP 403, a timeout.
Those refusals are not a flaw in those sites. They are bot protection declining a crawler it does not recognise. But a model that never reaches a page has no way to cite it. This check runs the same read on one address of yours.
No account, no email. Five checks a day from one connection.
What this check does
It requests the page once, with the same identity our crawler uses, and reports whether the server answered. It does not score the page, follow its links, or ask a model anything. That is why it answers in seconds and needs nothing from you.
The method and the full reading of the 300 agencies are public: see the index.