Accessibility statement

Last reviewed:

Standard applied

This site aims at WCAG 2.2 AA with the exceptions declared below. That is not a claim of full conformance, and the difference matters: four success criteria are outside what our automated gate can reach, and each is listed with the human control that covers it instead.

How we test

Every build runs an automated accessibility scan over every page the site publishes, with the WCAG 2.0, 2.1 and 2.2 A and AA rule sets selected, and a single violation fails the build. The same chain validates the markup, checks that no internal link points at a missing page, and refuses a page that loads anything from a third party.

The scan runs in a simulated browser, which computes no page layout. That is the reason for two of the four gaps below, and it is why a keyboard and composition walkthrough is run by a person against the deployed preview rather than against the build.

Known gaps

These are the criteria our automated gate cannot decide, why it cannot, and what covers them in its place. A check that cannot fail is not a check, so each one is named rather than left to the reader to discover.

1.4.3 Minimum contrast
the rule exists and is switched off explicitly in our scan configuration
Covered instead by: The theme generator refuses to emit a colour pair whose contrast has not been measured, so no unmeasured pair can enter the system. Its limit, said plainly rather than left to be assumed: it covers pairs of design tokens, and it does not cover text over images, the focus indicator (1.4.11), disabled states, or a surface composed through transparency. That last one is why no surface on this site uses transparency at all.
2.5.8 Target size (minimum)
the rule exists and is switched off explicitly in our scan configuration
Covered instead by: A design review, run by a named person against a written list of eight targets — the mobile menu button and its entries, the navigation links, the form fields, the send button, the footer links, the entity table source links and the hero link — each with its own floor and its method of measurement, checked at the 320px width where targets are smallest. It ran when the layout shell was built and runs again whenever a page adds new targets, and the measurements are recorded where the change is reviewed.
2.4.6 Headings and labels
the rule exists but carries no WCAG tag, so our tag selection does not run it
Covered instead by: Manual review of the heading hierarchy on every page the site publishes. A rule for this does exist in our scanner, but it is classed as best practice rather than carrying a WCAG tag, so the tag selection above does not run it. Selecting that one rule as well is a decision we have not taken yet, and until we do, the control is the review.
2.4.7 Focus visible
no rule is published for this criterion
Covered instead by: A keyboard walkthrough run by a person against the deployed site: every element that can take focus shows its focus ring, and none loses it. No rule is published for this criterion by any scanner we know of, and our scan environment computes no layout even if one existed, so no automated control is possible without driving a real browser. We say so rather than let the absence pass for coverage.

Contact

If something on this site is not usable for you, tell us and we will fix it. Say which page and what happened; you do not need to know why it failed.

Email: contact@gefin.uk