accessibility-review

/accessibility-review

If you see unfamiliar placeholders or need to check which tools are connected, see CONNECTORS.md.

Audit a design or page for WCAG 2.1 AA accessibility compliance.

Usage

/accessibility-review $ARGUMENTS

Audit for accessibility: @$1

WCAG 2.1 AA Quick Reference

Perceivable

• 1.1.1 Non-text content has alt text
• 1.3.1 Info and structure conveyed semantically
• 1.4.3 Contrast ratio >= 4.5:1 (normal text), >= 3:1 (large text)
• 1.4.11 Non-text contrast >= 3:1 (UI components, graphics)

Operable

• 2.1.1 All functionality available via keyboard
• 2.4.3 Logical focus order
• 2.4.7 Visible focus indicator
• 2.5.5 Touch target >= 44x44 CSS pixels

Understandable

• 3.2.1 Predictable on focus (no unexpected changes)
• 3.3.1 Error identification (describe the error)
• 3.3.2 Labels or instructions for inputs

Robust

• 4.1.2 Name, role, value for all UI components

Common Issues

1. Insufficient color contrast
2. Missing form labels
3. No keyboard access to interactive elements
4. Missing alt text on meaningful images
5. Focus traps in modals
6. Missing ARIA landmarks
7. Auto-playing media without controls
8. Time limits without extension options

Testing Approach

1. Automated scan (catches ~30% of issues)
2. Keyboard-only navigation
3. Screen reader testing (VoiceOver, NVDA)
4. Color contrast verification
5. Zoom to 200% — does layout break?

Output

## Accessibility Audit: [Design/Page Name]
**Standard:** WCAG 2.1 AA | **Date:** [Date

If Connectors Available

If ~~design tool is connected:

• Inspect color values, font sizes, and touch targets directly from Figma
• Check component ARIA roles and keyboard behavior in the design spec

If ~~project tracker is connected:

• Create tickets for each accessibility finding with severity and WCAG criterion
• Link findings to existing accessibility remediation epics

Tips

1. Start with contrast and keyboard — These catch the most common and impactful issues.
2. Test with real assistive technology — My audit is a great start, but manual testing with VoiceOver/NVDA catches things I can't.
3. Prioritize by impact — Fix issues that block users first, polish later.