docs
Contents
Concepts

Audits and agent readiness

What an audit checks, how it scores, and how agent readiness combines Webtick's checks with Vercel Is Agentic.

An audit crawls up to 25 public pages of the site and runs a fixed, versioned set of checks. Each check has a status (pass, fail, partial, not applicable, error), a severity and a weight, and names the page it applies to. Scores are computed per dimension; the methodology version is stored with every audit so results stay comparable.

Dimensions

DimensionIn the core scoreChecks
Technical SEOYesSuccess status, title length, meta description, same-site canonical, indexable, one H1, internal links resolve, robots.txt, XML sitemap
Structured dataYesJSON-LD is valid JSON, declares schema.org context and type, agrees with visible content; Organization or WebSite identity
Content accessibilityYesMain content without JavaScript, image alt attributes, declared language
Agent readinessYesrobots.txt permits AI crawlers, pages readable without JavaScript, a machine-readable summary on the home page
Optional conventionsNollms.txt, humans.txt
External benchmarkNoVercel Is Agentic score, shown beside ours

Agent readiness and Vercel Is Agentic

Agent readiness shows Webtick's own score and the latest Vercel Is Agentic score, each with its source and time. A combined score (the equal-weight mean) appears only when both exist and the Is Agentic scan is at most 90 days older than the audit. Findings from both sources are shown in one list, labelled by source.

Webtick reads Is Agentic's public report for the site. For a project audit, when no report exists yet, Webtick starts one scan; that publishes a public report of the site on is-agentic.com. A missing report is checked again after 10 minutes, 1 hour and 6 hours. Free public audits only look reports up and never start scans.

History

Project audits are kept as a history with score changes against the previous audit. Start a new audit from the project at any time.

Next: Business context