$npx -y skills add Owl-Listener/inclusive-design-skills --skill decision-documentationDocument accessibility decisions and the reasoning behind them so they survive team changes, redesigns, and time. Use when making accessibility tradeoffs, choosing between approaches, or when a decision needs to be recorded for future reference. Triggers on: accessibility decisio
| 1 | # Accessibility Decision Documentation |
| 2 | |
| 3 | Record accessibility decisions so that future teams understand not |
| 4 | just what was decided, but why — preventing the cycle where good |
| 5 | decisions get reversed because nobody remembers the reasoning. |
| 6 | |
| 7 | ## Why This Matters |
| 8 | |
| 9 | Accessibility barriers accumulate quietly through redesigns, framework |
| 10 | updates, staff turnover, and rushed deadlines. The most common cause |
| 11 | isn't malice or ignorance — it's a new team member changing something |
| 12 | without knowing why it was built that way. |
| 13 | |
| 14 | Documentation is the immune system. It protects good decisions from |
| 15 | being undone by people who weren't in the room. |
| 16 | |
| 17 | ## Decision Record Format |
| 18 | |
| 19 | For each significant accessibility decision, record: |
| 20 | |
| 21 | ### Title |
| 22 | A clear, searchable name. |
| 23 | "Form validation: inline errors instead of summary banner" |
| 24 | |
| 25 | ### Date and Author |
| 26 | When was this decided and by whom. |
| 27 | |
| 28 | ### Context |
| 29 | What situation prompted this decision? What were the constraints? |
| 30 | "During user testing, 3 out of 5 participants with cognitive |
| 31 | disabilities missed the error summary banner at the top of the page. |
| 32 | They corrected individual fields but didn't scroll up to see the |
| 33 | banner, so they didn't know the form hadn't submitted." |
| 34 | |
| 35 | ### Decision |
| 36 | What was decided, specifically. |
| 37 | "All form validation errors will appear inline directly below |
| 38 | the field that needs correction. The error summary banner is |
| 39 | removed. Errors are announced to screen readers via |
| 40 | aria-describedby on each field." |
| 41 | |
| 42 | ### Alternatives Considered |
| 43 | What other options were evaluated and why they were rejected. |
| 44 | "We considered keeping the summary banner alongside inline errors, |
| 45 | but testing showed it created confusion — users didn't know which |
| 46 | error indication to follow." |
| 47 | |
| 48 | ### Accessibility Impact |
| 49 | Who benefits from this decision and how. |
| 50 | "Benefits users with cognitive disabilities (don't need to |
| 51 | remember which fields had errors), screen reader users (errors |
| 52 | announced in context), and all users (faster error correction)." |
| 53 | |
| 54 | ### Risks and Tradeoffs |
| 55 | What might go wrong or what was sacrificed. |
| 56 | "If a form has many errors, the user sees them one at a time as |
| 57 | they move through fields, rather than all at once. For long forms, |
| 58 | this could mean more scrolling. We accept this tradeoff because |
| 59 | contextual errors have higher fix rates in testing." |
| 60 | |
| 61 | ### Evidence |
| 62 | What research, testing, or standards support this decision. |
| 63 | "User testing session 2024-03-15 (5 participants, 3 with cognitive |
| 64 | disabilities). WCAG 2.2 SC 3.3.1 (Error Identification)." |
| 65 | |
| 66 | ## When to Document |
| 67 | |
| 68 | Document a decision when: |
| 69 | - You chose between two or more accessibility approaches |
| 70 | - You made a tradeoff that reduces accessibility in one area |
| 71 | to improve it in another |
| 72 | - You decided NOT to fix something (and why) |
| 73 | - You implemented something differently from WCAG techniques |
| 74 | for a good reason |
| 75 | - A design pattern was chosen specifically for accessibility reasons |
| 76 | - User testing revealed an accessibility insight that changed direction |
| 77 | |
| 78 | ## Where to Store |
| 79 | |
| 80 | - In the codebase alongside the code (preferred — travels with |
| 81 | the implementation) |
| 82 | - In a shared decision log (wiki, Notion, Confluence) |
| 83 | - In design system documentation (for system-level decisions) |
| 84 | - Never only in someone's head or in a Slack thread |
| 85 | |
| 86 | ## Assessment Questions |
| 87 | |
| 88 | 1. Are accessibility decisions documented with reasoning? |
| 89 | 2. Can a new team member find out why something was built a |
| 90 | specific way? |
| 91 | 3. Are tradeoffs explicitly recorded? |
| 92 | 4. Is documentation stored where it won't be lost? |