What Shipped, and When

Partial Audits Count — a crawl that stops at its page budget is marked partial, and never compared to a full one as if the site regressed. Deltas count only shared pages; partial cards point back to your last full audit.

Four Severity Levels — Critical, Warning, Opportunity, and Information, so the worklist reads as a plan instead of a pile.

Lens switching is now instant — results are cached per crawl, so moving between Overview, Issues, and Tasks no longer reloads from scratch.

GEO Signals — the Answer brain checks AI-crawler access, llms.txt, JS-only content risk, and answer readiness, so one URL is judged for both search and assistants.

Signal Intelligence — the decision core now surfaces where the nine brains disagree, and says what it could not measure rather than implying a clean result.

Findings group by page template — “your article template does this on 349 pages,” not 349 addresses.

GSC & GA4 Integrations — connect read-only to fold real demand and behaviour into every decision, with Core Web Vitals from CrUX and PageSpeed.

Traffic at Risk — crawl, Search, and Analytics joined on one page key to flag valuable pages quietly losing ground.

Evidence panels on every finding — the crawl, intent, and demand proof behind a verdict, not just a score.

White-Label Reports — the executive decision first, evidence underneath, shareable as a link or a report-only client account.

Team Roles — SEO Manager, Editor, and report-only Client, each scoped server-side.

Crawl budget explorer explains exactly why a crawl stopped where it did.