Site audit reads indexing and snippet signals

The site audit checks more of what decides whether a page can appear in search and AI answers.

• Each page is checked for noindex, nosnippet, max-snippet, data-nosnippet, redirect chains, canonical targets, hreflang return links and login walls. The googlebot meta and X-Robots-Tag are read too.
• robots.txt is worked out for each crawler: AI search, Googlebot and Bingbot, user-triggered fetchers and AI training. Only blocks of search crawlers count as problems.
• A random missing path is requested once to find soft 404s.
• Chinese and Japanese pages are counted by words, not by spaces.
• Each audit is compared with the previous one. New critical issues and new robots.txt blocks send an in-app notification, and the weekly digest lists the week's new issues.
• Subscribed projects can choose a daily audit of up to 200 pages under Settings → Sites. The weekly audit stays free for everyone.

AI readiness rules are now version 1.7.0: blocking AI search crawlers costs score, nosnippet fails F04, and llms.txt no longer adds points. The first check after the update starts a new baseline. The growth agent no longer recommends FAQ schema.