ai visibility audit.

Eight things, in one pass: meta signals (title, description, canonical, robots, Open Graph), the extracted content a crawler actually reads, per-bot readiness for the major AI crawlers, structured data (schema.org), robots.txt AI-crawler directives, your XML sitemap, llms.txt, and emerging agent-readiness signals.

It analyses static HTML only, with no JavaScript rendering. That is deliberate: it reflects what most AI crawlers (ChatGPT, Claude, Perplexity) actually receive, since they do not run your JavaScript. Only Google and Gemini render the full page, and the tool is honest where it cannot replicate that.

What it reads straight off your page - HTTP status, meta tags, canonical, robots.txt, schema, sitemap, llms.txt - is exact, pulled directly from the response. The per-bot view (how ChatGPT, Claude and the rest process your page) is a best-effort model of behaviour the AI companies mostly do not document, so treat it as indicative, not gospel; each bot's tab flags where a step is inferred. And since it reads static HTML only, anything rendered with JavaScript is reported as missing rather than executed. When in doubt, the word count and the raw extracted text are the honest ground truth.

Mostly, what helps an LLM find you is technical SEO hygiene done well. A few things really do shift in AI search, and where they do this tool flags them, but it will not pretend a robots line or a schema block is a growth hack. Where the evidence is thin (llms.txt, schema as an LLM signal), it says so instead of scoring you on it.