Skip to content
Footnoted
Register

Check. Fix. Check again.

Enter a page. Get a score and your next steps. Re-check after you make changes.

  1. 1

    Enter a URL

    Any page you want AI assistants to use: your homepage, a product or service page, a guide. No account is needed for the free check.

  2. 2

    We read it like an AI crawler

    We fetch the HTML your server sends, without running JavaScript — many AI crawlers do not run it either. We also read your robots.txt, look for an XML sitemap and an llms.txt file, and time the response.

  3. 3

    We score what matters

    More than twenty checks across six areas — crawler access, content structure, structured data, metadata, trust and freshness, and AI discovery — add up to a score out of 100. Checks that decide whether a page can be used at all weigh the most.

  4. 4

    You get the fixes

    Every check says what we found and how to fix it, and your top fixes are ranked by how much each one is costing you. Download the report as a PDF for your developer or share the link.

  5. 5

    Fix and re-check

    Make the changes, then scan again to see the score move. With a free account every report is kept on your dashboard.

Going further

How the extensive audit works

The extensive audit crawls up to 20 pages of your site — from your sitemap first, then from links — and scores each one, so you can see which problems repeat across templates. Our crawler obeys your robots.txt for the pages it picks.

It then tests what AI actually says. We work out what your site offers and write 5 questions a potential customer might ask before they know your brand. Each one goes to Claude with live web search on, without mentioning your name, and we record whether your site is cited, mentioned or missing, and which sites are cited instead. We also ask Claude what it knows about your brand from its training data alone.

Then it turns every issue into a fix. Each one is ranked from critical to optional and lists the pages it affects, why it matters, the steps, a ready-made snippet (robots.txt lines, JSON-LD, meta tags) and, when we can tell what your site is built with, where to click on that platform.

Every fix also comes with a prompt to paste into your AI assistant — ChatGPT, Claude, Gemini, Copilot or Cursor — carrying your site, the issue and our snippet, and asking the assistant to ask you rather than invent anything it does not know. One extra prompt walks you through all the fixes in order. Claude also drafts new titles, descriptions and question-led sections for your weakest pages, from their own text, with [bracketed] gaps where a fact is missing.

Finally you get a prioritised action plan, a starter llms.txt built from your pages, Organization schema ready to paste into your homepage, and the whole fix pack as a download.

What a score can and can't tell you
  • A high score is not a guarantee of citations. It means your page did well on the signals we checked. Whether they choose it still depends on your content, your reputation and the question asked.
  • AI answers vary. Ask the same question twice and you can get different sources. The extensive audit's AI test is a snapshot, not a ranking.
  • llms.txt is an emerging convention. We check for it and generate one in the extensive audit, but the major assistants have not confirmed they use it, so it carries little weight in the score.
  • We see what our crawler sees. If a firewall treats AI crawlers differently from ours, results can differ. When our crawler is blocked, the report says so.
  • Assistants change. The crawlers and signals that matter change over time. We update the checks as vendors publish new crawler names and guidance.

Try it on your site.

The free check takes seconds.

Check my site → See every check