Skip to main content

The engine

Ility scans each page in a real Chromium browser, so it sees the page as your visitors do, after scripts have run. The checks themselves come from axe-core, the open-source accessibility engine from Deque Systems.

Which rules run

Every scan runs two kinds of rules.

WCAG rules

Rules mapped to a WCAG success criterion at or below the surface’s conformance target. With a 2.1-AA target, that is every rule for a level A or AA criterion in WCAG 2.0 or 2.1.

Best-practice rules

Rules that improve accessibility but don’t map to a WCAG criterion, such as having exactly one main landmark. They run on every scan, whatever the target. Their issues are labeled best practice and don’t count against your WCAG conformance.
Ility only runs stable rules. Rules the engine marks as experimental or deprecated are left out, so every finding is one you can defend in an audit.

Automated coverage

The Automated coverage card on a surface’s Overview shows how much of the surface and of WCAG the last scans reached:
  • Pages scanned: how many of the pages Ility knows about have been scanned.
  • WCAG criteria evaluated: the share of success criteria in your target that at least one automated check covers. It shows once the surface has an explicit WCAG target.
  • Passed, Failed, and Needs review: how the individual checks came out. A check needs review when the engine found something it can’t judge on its own, such as text over a background image.

What automation can’t catch

Automated testing finds a large share of common accessibility problems, but not all of them. Many WCAG criteria need a person to judge, for example whether alt text actually describes the image or whether the reading order makes sense. A clean scan means no automated check failed. It doesn’t on its own mean a site conforms to WCAG.