Seven facts about your home page and root files that decide whether ChatGPT, Perplexity and Google's AI surfaces can fetch, parse and quote you — each with the evidence, none with a made-up grade. Takes a few seconds; no card, no account.
Checks: server-rendered text · direct-answer structure · title and description · JSON-LD · AI crawlers in robots.txt · llms.txt · sitemap
Most answer engines fetch the HTML and parse it; the ones that render JavaScript do so unreliably. Text that exists only after scripts run is text that may not exist for them.
An answer engine lifts the sentence or table that answers the question. A page structured as heading → direct answer → detail is the shape that gets lifted.
GPTBot, PerplexityBot and the rest obey robots.txt. A wildcard block written for scrapers in 2019 quietly removes you from the assistants people use in 2026.
This check reads the home page and three root files once and caches the result for an hour. It does not crawl the site, store the HTML, or measure whether you are cited — that is the measurement.