| name | design-accessibility |
| description | WCAG AA and ARIA best practices — screen readers, keyboard navigation, focus management. Use when building any user-facing interface or reviewing accessibility compliance. |
| allowed-tools | ["Bash","Read","Write","Edit","Grep","Glob","Agent"] |
| metadata | {"category":"frontend","tags":["accessibility","wcag","aria","a11y"],"status":"ready","version":5} |
Principles
- Use semantic HTML first — ARIA is a last resort, not a first tool
- Proper heading hierarchy — one h1 per page, never skip levels
- Alt text on all meaningful images
- Maintain 4.5:1 minimum color contrast ratio
- Use
<button> for actions, <a> for navigation — never a styled <div>
Rules
See rules index for detailed patterns.
Examples
Positive Trigger
User: "Audit this form for WCAG AA issues and keyboard traps."
Expected behavior: Use design-accessibility guidance, follow its workflow, and return actionable output.
Non-Trigger
User: "Optimize SQL indexes for this analytics query."
Expected behavior: Do not prioritize design-accessibility; choose a more relevant skill or proceed without it.
Troubleshooting
Skill Does Not Trigger
- Error: The skill is not selected when expected.
- Cause: Request wording does not clearly match the description trigger conditions.
- Solution: Rephrase with explicit domain/task keywords from the description and retry.
Guidance Conflicts With Another Skill
- Error: Instructions from multiple skills conflict in one task.
- Cause: Overlapping scope across loaded skills.
- Solution: State which skill is authoritative for the current step and apply that workflow first.
Output Is Too Generic
- Error: Result lacks concrete, actionable detail.
- Cause: Task input omitted context, constraints, or target format.
- Solution: Add specific constraints (environment, scope, format, success criteria) and rerun.
Workflow
- Identify whether the request clearly matches
design-accessibility scope and triggers.
- Apply the skill rules and referenced guidance to produce a concrete result.
- Validate output quality against constraints; if gaps remain, refine once with explicit assumptions.
