What the current beta can inspect
The beta can inspect selected response status and security headers, HTTPS, canonical tags, viewport configuration, page title, meta description, headings, and basic content signals on the submitted public page.
robots.txt and sitemap checks use separate public requests. A discovered file that is unavailable, blocked, invalid, or unsafe is reported as unavailable rather than treated as a passing measurement.
The beta reviews public robots.txt rules for selected AI crawler user agents. This can identify an apparent allow, block, or unknown state for those directives.
A robots.txt result cannot prove that an AI provider has crawled, indexed, trained on, cited, or will surface a page. Product names are used only to identify public crawler user agents and do not imply endorsement or affiliation.
Core Web Vitals are shown only when the configured PageSpeed integration returns a measurement. When it does not, the report labels performance as unmeasured and excludes it from the overall weighted score.
HTML heuristics may still identify possible implementation risks, such as a lazy-loaded first image or render-blocking scripts, but those observations are not substitutes for measured field or lab performance data.
