Figures you can quote: data as of, final, one window per page, and robots.txt by day
Every answer now says when it was computed and whether it is final, each page uses one time window or says plainly where Search Console's 28 days apply, URL pages pick the canonical form, Crawl Join says when Googlebot last fetched each page, and robots.txt responses are classed by day and bot the way Google reads them.
Audits.com re-derives every figure before quoting it to a client. These changes make the figures quotable as they stand.
What changed
- Data as of, and final. Every per-site answer from the API and the MCP server, and its receipt, says when it was computed and whether its window has closed. A window that closed at least 30 minutes ago with nothing still buffering is final and will not move; a window that runs to now is provisional, and the answer says so.
- One definition per term.
GET /public/v1/definitionsand the MCP toolget_definitionsgive one definition for every metric and status the reports use, from verification statuses to z-score, so a quoted figure can carry its meaning. - One window per page. Search Performance, Search reconcile and the AI funnel now say which blocks follow the window you selected and which are Search Console's trailing 28 days, in the response and beside the table. Page Importance reports one window, 90 days, with the days actually covered beside it.
- Which form is the page. URL detail shows both
/aboutand/about/and picks the canonical one from the sitemap, else the form that answers 200, and compares Search Console and analytics figures against it. - Last Googlebot fetch per page. Crawl Join rows carry the last verified Googlebot request and the days since; a page without one reads "not requested by verified Googlebot in these N days", never "never crawled".
- robots.txt responses by day and bot. A new report and API endpoint class each day's robots.txt fetches the way Google does (file, no file, unsuccessful, redirect) and list every unsuccessful Googlebot fetch with Googlebot's requests in the minute either side, so a 5xx window can be seen to stop crawling. MCP 1.24.0 adds
get_robots_responses; the CLI (2.4.0) addsseo robots-responsesanddefinitions.
← Every page now says what your feed can and can't measureBots are classed by what they do: ads crawlers, SEO tools and Applebot as search →
← All changelog entries