How scanning works

What is tested

• Every page is loaded in a real browser at desktop size (1366 × 900) and phone size (390 × 844). Findings are merged, and the report says at which size each one appeared.
• Pages are tested against WCAG 2.2 levels A and AA with axe-core, plus AccessibilityBrief's own checks: page language, skip links, visible focus, vague link text, heading order, several H1s, autoplaying media and touch target size on phones.
• Best-practice suggestions are listed separately and never lower the score.
• PDF documents linked from scanned pages are checked too (tags, text, title, language, bookmarks, restrictions). Their findings are listed but are not part of the score.

How pages are found

The scan starts at the address you gave, reads your sitemap and follows links on the same site. Addresses are tidied (tracking parameters such as utm_*, fbclid and gclid are removed) so the same page is not scanned twice. On rescans, pages whose HTML has not changed reuse the previous result.

How the scanner behaves

• It identifies itself as AccessibilityBriefBot. See the bot page for details.
• It follows your robots.txt rules for AccessibilityBriefBot and for all bots. If robots.txt cannot be read because of a server error, it does not scan.
• It requests at most two pages at a time from one site, with a short pause between requests.
• It never logs in or submits forms, unless you set up scanning behind a login.
• It only connects to public internet addresses.

If a firewall or CDN blocks the scanner, see Bot protection and firewalls.