gpt-taste
gpt taste
a11y
Production-grade accessibility skill for WCAG 2.2 AA compliance. Covers auditing, remediation, component authoring, and validation workflows. Auto-invoked for UI implementation, a11y fixes, and accessibility testing.
Full skill instructions
Accessibility implementation guide aligned with WCAG 2.2 Level AA, WAI-ARIA 1.2, and WCAG2ICT.
| Standard | Scope | Normative Source |
|---|---|---|
| WCAG 2.2 | Web content | https://www.w3.org/TR/WCAG22/ |
| WAI-ARIA 1.2 | Widget semantics | https://www.w3.org/TR/wai-aria-1.2/ |
| ARIA APG | Authoring patterns | https://www.w3.org/WAI/ARIA/apg/ |
| WCAG2ICT | Non-web ICT | https://www.w3.org/TR/wcag2ict-22/ |
| EN 301 549 | EU procurement | ETSI EN 301 549 V3.2.1 |
| SC | Level | Requirement | Implementation |
|---|---|---|---|
| 2.4.11 | AA | Focus Not Obscured (Minimum) | Ensure focused element is at least partially visible |
| 2.4.13 | AAA | Focus Appearance | Focus indicator area ≥ 2px perimeter, 3:1 contrast |
| 2.5.7 | A | Dragging Movements | Provide single-pointer alternative to drag operations |
| 2.5.8 | AA | Target Size (Minimum) | 24×24 CSS pixels minimum |
| 3.2.6 | A | Consistent Help | Place help mechanisms in same relative location |
| 3.3.7 | A | Redundant Entry | Auto-populate previously entered information |
| 3.3.8 | AA | Accessible Authentication | No cognitive function test for login |
| SC | Requirement | Implementation | Test |
|---|---|---|---|
| 1.1.1 | Non-text content | alt on images, aria-label on icon buttons | img[alt], button[aria-label] |
| 1.3.1 | Info and relationships | Semantic HTML, no <div> for structure | Landmark audit |
| 1.4.3 | Contrast (minimum) | 4.5:1 text, 3:1 large text | axe color-contrast |
| 1.4.11 | Non-text contrast | 3:1 UI components/graphics | Manual inspection |
| SC | Requirement | Implementation | Test |
|---|---|---|---|
| 2.1.1 | Keyboard | All functions keyboard-accessible | Tab traversal |
| 2.4.3 | Focus order | Logical DOM sequence | Visual focus path |
| 2.4.7 | Focus visible | 2px+ visible indicator | focus-visible:ring-2 |
| 2.5.8 | Target size | Minimum 24x24 CSS pixels | min-w-6 min-h-6 |
| SC | Requirement | Implementation | Test |
|---|---|---|---|
| 3.2.1 | On focus | No context change on focus | Focus does not submit |
| 3.3.1 | Error identification | Text description, not color alone | role="alert" present |
| SC | Requirement | Implementation | Test |
|---|---|---|---|
| 4.1.2 | Name, role, value | Accessible name on all controls | axe button-name |
// REQUIRED: Landmark structure
<header role="banner">...</header>
<nav aria-label="Main">...</nav>
<main role="main">...</main>
<aside role="complementary">...</aside>
<footer role="contentinfo">...</footer>
// PROHIBITED: Div-based structure
<div class="header"> // SC 1.3.1 violation
<div onClick={...}> // SC 2.1.1, 4.1.2 violation
// BAD
<button onClick={handleDelete}><Trash2 /></button>
// GOOD
<button type="button" aria-label="Delete" onClick={handleDelete}>
<Trash2 aria-hidden="true" />
</button>
// BAD
<div className="btn" onClick={handleClick}>Save</div>
// GOOD
<button type="button" onClick={handleClick}>Save</button>
// BAD
<input placeholder="Search..." />
// GOOD
<label htmlFor="search">Search</label>
<input id="search" type="search" />
// OR
<input type="search" aria-label="Search files" />
// BAD
<input className={error ? "border-red-500" : ""} />
// GOOD
<input aria-invalid={!!error} aria-describedby="error-msg" />
{error && <span id="error-msg" role="alert">{error}</span>}
// BAD
<img src="/logo.png" />
// GOOD: Informative
<img src="/logo.png" alt="Company logo" />
// GOOD: Decorative
<img src="/decoration.png" alt="" role="presentation" />
/* BAD: ~2.5:1 ratio */
.muted { color: oklch(75% 0 0); }
/* GOOD: 4.5:1+ ratio */
.muted { color: oklch(45% 0 0); }
Never rely on color alone to convey information (SC 1.4.1):
// BAD: Color-only status
<span className={status === "error" ? "text-red-500" : "text-green-500"}>
{status}
</span>
// GOOD: Color + icon + text
<span className={status === "error" ? "text-red-500" : "text-green-500"}>
{status === "error" ? <AlertCircle aria-hidden /> : <CheckCircle aria-hidden />}
{status === "error" ? "Error" : "Success"}
</span>
Testing tools:
// BAD
<button className="outline-none">Action</button>
// GOOD
<button className="focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2">
Action
</button>
// BAD
<h1>Title</h1>
<h3>Subsection</h3> // h2 skipped
// GOOD
<h1>Title</h1>
<h2>Subsection</h2>
// BAD: 16x16
<button className="p-0.5"><X size={12} /></button>
// GOOD: 24x24 minimum
<button className="p-2 min-w-6 min-h-6"><X size={16} /></button>
// BAD
<video src="/demo.mp4" autoPlay />
// GOOD
<video src="/demo.mp4" autoPlay muted />
<ul role="tree" aria-label="File explorer">
<li
role="treeitem"
aria-expanded={isExpanded}
aria-selected={isSelected}
aria-level={level}
aria-setsize={siblingCount}
aria-posinset={position}
tabIndex={isFocused ? 0 : -1} // Roving tabindex
>
<span>{name}</span>
{hasChildren && (
<ul role="group">{children}</ul>
)}
</li>
</ul>
Keyboard: ↓↑ navigate, → expand/child, ← collapse/parent, Enter activate, Space toggle, Home/End boundaries
<div
role="dialog"
aria-modal="true"
aria-labelledby="dialog-title"
tabIndex={-1}
>
<h2 id="dialog-title">Title</h2>
{/* Focus trap: Tab cycles within */}
{/* Escape: closes dialog */}
{/* On close: restore focus to trigger */}
</div>
<div role="alert" aria-live="polite" aria-atomic="true">
{message}
</div>
// Error: aria-live="assertive"
// Only focused item has tabIndex={0}
{items.map((item, i) => (
<button
key={item.id}
tabIndex={focusedIndex === i ? 0 : -1}
onKeyDown={(e) => handleArrowKeys(e, i)}
/>
))}
// Store trigger before modal opens
const triggerRef = useRef<HTMLElement>(null)
const openModal = (e) => { triggerRef.current = e.currentTarget; setOpen(true) }
const closeModal = () => { setOpen(false); triggerRef.current?.focus() }
@media (prefers-reduced-motion: reduce) {
*, *::before, *::after {
animation-duration: 0.01ms !important;
transition-duration: 0.01ms !important;
}
}
# axe-core
npx @axe-core/cli <url> --tags wcag2a,wcag2aa,wcag22aa --exit
# Lighthouse
npx lighthouse <url> --only-categories=accessibility --output=json
# pa11y
npx pa11y <url> --standard WCAG2AA
prefers-reduced-motion, verify complianceNVDA (Windows):
Insert+Space to toggle focus modeInsert+Down to read allH to navigate headings, D for landmarksVoiceOver (macOS):
Cmd+F5 to enableVO+A to read allVO+U to open rotor (headings, links, landmarks)Mobile (TalkBack/VoiceOver):
## Component: [Name]
### WCAG Compliance
- [ ] SC 1.3.1: Semantic structure
- [ ] SC 2.1.1: Keyboard operable
- [ ] SC 2.4.7: Focus visible
- [ ] SC 4.1.2: Accessible name
### Screen Reader
Expected: "[Role]: [Name], [State]"
gpt taste
Premium brand-kit image generation for high-end identity systems, logo concepts, and visual-world presentations.
Premium website design-to-code skill that generates visual references first, then builds faithful frontend implementations.
Premium frontend image direction for landing pages and marketing sites, generating one distinct horizontal image per section.
Premium mobile app screen concepts and flows with clean hierarchy, strong imagery, and distinctive art direction.
Emil Kowalski's philosophy on UI polish, component design, and the invisible details that make software feel good.
Frontend design and iteration for production-grade interfaces, from concept to ship.
Reverse-engineer design tokens from public websites into starter token files for your project.
Builds an animation from scratch by making the decisions in the order that determines whether it feels right.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.
Chat-to-edit AI studio for product photos, portraits, branding, and video.
Generate your dream room in seconds.
Redesign your interior in seconds using AI
Transform rooms with AI-generated designs.
Transform Your Space with AI-Powered Japanese Interior Designs by Jife
AI-powered wireframes for non-designers.
Revolutionize Property Viewing with AI-Generated Floor Plans and Virtual Tours
Transform sketches into stunning architectural renders with AI.