$npx -y skills add Owl-Listener/inclusive-design-skills --skill handoffGenerate an accessibility decision handoff for engineering. Chains: decision-documentation, compliance-mapping, accessibility-testing-strategy. Use when a design is ready for implementation and the engineering team needs clear accessibility specifications.
| 1 | # Accessibility Decision Handoff |
| 2 | |
| 3 | Package accessibility decisions into a clear specification that |
| 4 | engineers can implement against — because "make it accessible" |
| 5 | is not a specification. |
| 6 | |
| 7 | ## Process |
| 8 | |
| 9 | ### Step 1: Compile Decisions |
| 10 | Using **decision-documentation**, gather all accessibility decisions |
| 11 | for the feature: |
| 12 | - Interaction patterns and keyboard behaviour |
| 13 | - Screen reader announcements and ARIA requirements |
| 14 | - Visual specifications (contrast, colour independence, motion) |
| 15 | - Content requirements (reading level, alt text, link text) |
| 16 | - Adaptive behaviour (preferences, zoom, density) |
| 17 | |
| 18 | ### Step 2: Map to Implementation |
| 19 | Using **compliance-mapping**, translate each design decision into |
| 20 | specific implementation requirements: |
| 21 | |
| 22 | For each interactive element specify: |
| 23 | - HTML element or ARIA role |
| 24 | - Keyboard behaviour (which keys do what) |
| 25 | - Focus management (where focus goes after actions) |
| 26 | - Screen reader name, role, and state announcements |
| 27 | - Visual states (default, hover, focus, active, disabled, error) |
| 28 | |
| 29 | For each content element specify: |
| 30 | - Heading level |
| 31 | - Alt text (exact text or strategy) |
| 32 | - Link text |
| 33 | - Language attributes if multilingual |
| 34 | |
| 35 | For each adaptive behaviour specify: |
| 36 | - Media query and expected response |
| 37 | - Preference detected and what changes |
| 38 | - Breakpoints and reflow behaviour |
| 39 | |
| 40 | ### Step 3: Define Test Criteria |
| 41 | Using **accessibility-testing-strategy**, provide specific test |
| 42 | cases for each requirement: |
| 43 | - Keyboard test: "Tab to the submit button, press Enter, |
| 44 | verify focus moves to the confirmation message" |
| 45 | - Screen reader test: "Navigate to the error, verify it |
| 46 | announces: 'Error: email address is required'" |
| 47 | - Visual test: "Verify contrast ratio of body text is at |
| 48 | least 4.5:1 in both light and dark modes" |
| 49 | - Zoom test: "At 200% zoom, verify the form reflows to |
| 50 | single column with no horizontal scroll" |
| 51 | |
| 52 | ### Step 4: Flag Risks and Dependencies |
| 53 | Note anything engineers need to be aware of: |
| 54 | - Components that need custom keyboard handling |
| 55 | - Places where framework defaults aren't accessible |
| 56 | - Third-party components that may need wrapping or replacing |
| 57 | - Areas where automated testing won't catch the issue |
| 58 | |
| 59 | ## Output |
| 60 | |
| 61 | Deliver a handoff document containing: |
| 62 | |
| 63 | 1. **Accessibility specification** — element-by-element requirements |
| 64 | with HTML, ARIA, keyboard, and visual specifications |
| 65 | 2. **Test cases** — specific, executable test scenarios for keyboard, |
| 66 | screen reader, visual, and zoom testing |
| 67 | 3. **WCAG mapping** — which success criteria each requirement |
| 68 | addresses |
| 69 | 4. **Risk register** — known challenges and recommended approaches |
| 70 | 5. **Definition of done** — the specific accessibility criteria |
| 71 | that must pass before the feature ships |