Policy

Accessibility Statement

What is built into this site, what we have tested, and what we have not. Written to be checkable rather than reassuring.

Last revised 25 September 2026

The target

WCAG 2.2 Level AA. Accessibility was part of building this rather than a pass at the end, which is why the list below is specific.

What is in place

  • Semantic HTML, one H1 per page, headings in order.
  • Full keyboard operability, a visible focus indicator on keyboard navigation only, and a skip-to-content link.
  • Every hover behaviour has a focus equivalent. No information exists in hover alone.
  • Text contrast at or above 4.5:1, interface components at or above 3:1.
  • Disclosures use native details and summary elements. Tabs and the audience switch implement the documented ARIA patterns, including arrow-key navigation.
  • Form fields have persistent visible labels. Errors are linked programmatically and never signalled by colour alone.
  • Animation and smooth scrolling are removed entirely under prefers-reduced-motion, not merely shortened.
  • The 3D object is decorative. Everything it represents exists as real text beside it, and it is replaced by a static image under reduced motion, on small screens, or without WebGL.
  • The scrollbar is never hidden, so scroll position stays visible on a long page.
  • Interactive targets are at least 44 by 44 pixels.
  • Inner scroll regions still scroll normally despite the custom wheel handling.
  • Keyboard-driven scrolling to a focused element is instant rather than animated, so focus never appears to drift.

What we have not verified

Automated tooling catches roughly a third of accessibility problems. Full conformance cannot be claimed from it, and we have not completed manual screen-reader testing or an independent expert audit. Treat this as "no known violations", not as certified conformance.

Gaps we are aware of:

  • No independent audit has been carried out.
  • Screen-reader behaviour has not been verified across NVDA, JAWS and VoiceOver.
  • The custom wheel smoothing has been reasoned about carefully for keyboard and inner-scroll behaviour, but not tested with every assistive setup.
  • Contrast has been designed to the ratios above but not measured on every state of every component.

Telling us about a barrier

Email support@craftitstudios.com with the page and what stopped you. We acknowledge within 2 working days and will give you either a fix timeline or another way to get what you needed meanwhile. If something here is stopping you dealing with us, that is our problem to fix, not yours to work around.