Accessibility
This skill should be used when building or reviewing UI for accessibility — keyboard navigation, screen readers, ARIA, color contrast, focus states, accessible forms, or WCAG compliance. Trigger phrases include "make it accessible", "a11y", "accessibility", "WCAG", "ARIA", "keyboard navigation", "screen reader", "color contrast", "focus states", "accessible form", "is this compliant", "EAA", "ADA". It targets WCAG 2.2 AA with concrete component rules and an axe-core test step.From its SKILL.md
npx -y skills add MartinOlivero/saas-builder --skill accessibilityAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
3.9 KB, 882 tokens by cl100k_base, as published. Nobody here has run it
Accessibility
This skill makes interfaces usable by everyone — keyboard-only users, screen-reader users, low-vision users — and meets the legal bar (the EU Accessibility Act and ADA Title II both reference WCAG). Accessibility is not a nice-to-have; it's increasingly a legal requirement and always a larger addressable market.
Analogy: a curb cut was built for wheelchairs, but everyone with a stroller, suitcase, or bike uses it. Accessible UI is the same — built for those who need it, better for everyone.
Dependencies
This skill is stronger when Vercel's web-design-guidelines skill (part of vercel-labs/agent-skills, ~27k⭐) is installed — it ships the ARIA/focus/keyboard checklist. Delegate the checklist to it when present; always pair with the axe-core test step below, which catches runtime violations a static checklist can't. Never block on a missing dependency — the rules below are the embedded fallback.
The target
WCAG 2.2 Level AA — the bar referenced by the EU EAA, ADA Title II, and Section 508. Automated tools catch only ~30-40% of issues, so combine automated tests with manual checks.
Discovery (max 3 questions, only if unknown)
- Are you building new components or auditing existing UI?
- Any specific complaint (keyboard trap, unreadable contrast, screen reader silent)?
- Is there a CI pipeline where an a11y test could run?
Step 1 — Semantic HTML first
Use native elements before ARIA: <button>, <nav>, <main>, <label>, <a>. "No ARIA is better than bad ARIA" — reach for ARIA only when no native element fits. Maintain a logical heading hierarchy (one <h1>, no skipped levels) and add a "skip to content" link.
Step 2 — Keyboard & focus
- Every interactive element reachable and operable by keyboard alone (Tab / Shift-Tab / Enter / Space / Esc). Test by unplugging the mouse.
- Never remove focus outlines without a replacement. Style
:focus-visibleinstead ofoutline: none. - No keyboard traps — focus must be able to leave any widget (modals: trap within while open, release on close).
Step 3 — Forms
- Every input has a programmatically associated
<label>(oraria-label). - Link errors with
aria-describedby+ setaria-invalidon the field. - Don't rely on color alone to signal errors — add text/icon.
Step 4 — Color & targets
- Contrast ≥ 4.5:1 for normal text, ≥ 3:1 for large text and UI components/focus indicators.
- Touch targets ≥ 24×24px (WCAG 2.2 SC 2.5.8); 44×44px recommended for primary actions.
Step 5 — Dynamic content & motion
- Announce async updates (toasts, validation, loading) with an
aria-live="polite"region. - Honor
prefers-reduced-motion— gate non-essential animation behind it (also a WCAG 2.3.3 requirement). Coordinate with theui-designmotion settings.
Step 6 — Test it
- Add
jest-axeand assert per component:expect(await axe(container)).toHaveNoViolations(). Run it in CI. - Manually test one full flow with a real screen reader (VoiceOver on macOS, NVDA on Windows) — automated tools can't judge whether the announced experience makes sense.
Output
Deliver: the accessible components (semantic markup, labels, focus handling), the contrast/target fixes, an aria-live region for dynamic status, and a jest-axe test wired into CI.
Reference
w3c/wcag (WCAG 2.2), dequelabs/axe-core (~7k⭐), nickcolley/jest-axe, vercel-labs/agent-skills web-design-guidelines.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.
Gives 2 of the 12 instructions most accessibility skills give in 882 tokens
Counted across 360 of the 378 authors here whose files we hold, read 2026-09-06
- Respect prefers-reduced-motionin 49 of 360, across 41 files
- Add alt text to every imagein 47 of 360, across 40 files
- Make every interactive element keyboard operablehere, and in 46 of 360, across 37 files
- Keep focus indicators visiblein 42 of 360, across 37 files
- Associate every form input with a labelin 42 of 360, across 35 files
- Ensure 4.5:1 contrast for normal texthere, and in 38 of 360, across 32 files
- Trap focus in modals and restore it on closein 37 of 360, across 25 files
- Prefer native HTML elements over ARIAin 37 of 360, across 33 files
- Add aria-label to icon-only buttonsin 24 of 360, across 13 files
- Announce dynamic content updates with aria-livein 24 of 360, across 16 files
- Use ARIA only when native HTML is insufficientin 23 of 360, across 15 files
- Meet WCAG AA contrast ratiosin 23 of 360
Said here and by no other author read
- Wire jest-axe assertions into CI
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.