Site auditMEASURED

What does your site tell AI bots? Read it the way they do.

Plumb fetches your robots.txt, llms.txt, declared sitemaps and homepage, grades the rules with the same method as the robots.txt leaderboard, and sets the result next to publishers in your region.

Free, no account. Five audits a minute and fifty a day per network. Results are kept for an hour and the link to them can be shared.

01What is checked

A handful of requests, nothing hidden

  • robots.txt, bot by bot

    Twenty AI crawler tokens, grouped into training, AI search and user-requested fetches, each graded blocked, partial or allowed. The rules are the leaderboard's, written out in its methodology.

  • llms.txt and sitemaps

    Whether /llms.txt answers with a text file, and whether up to three of the sitemaps your robots.txt declares answer with a sitemap. With none declared, /sitemap.xml is tried. Sitemaps on another domain are listed, not fetched.

  • Homepage directives

    One request for the homepage, read for the X-Robots-Tag header and robots meta tags, including bot-specific ones and the non-standard noai and noimageai.

  • How Plumb asks

    As PlumbAudit/1.0 with a link back here, never posing as a browser or another bot. Redirects are followed only within your own domain, files are read up to a size cap, and the whole audit stops after 15 seconds. Private and reserved network addresses are refused.

  • What this cannot tell you

    What bots actually did. A robots.txt is a request. Whether GPTBot, or an impostor using its name, read your pages is in your edge log, which is what Plumb measures once a site is connected.