$npx -y skills add mgifford/accessibility-skills --skill plannerThe Planner identifies and categorizes accessibility findings, making them actionable for human professionals while maximizing automation where possible.
| 1 | # Planner Workflow |
| 2 | |
| 3 | The Planner identifies and categorizes accessibility findings, making them actionable for human professionals while maximizing automation where possible. |
| 4 | |
| 5 | ## Purpose |
| 6 | |
| 7 | The Planner's role is to: |
| 8 | |
| 9 | 1. **Map findings to ARRM tasks**: Identify which ARRM professional roles own responsibility for the finding |
| 10 | 2. **Distinguish automation**: Separate clear-rule issues from judgement calls |
| 11 | 3. **Identify perspectives**: Consider impacts on different user groups |
| 12 | 4. **Generate questions**: Ask role-specific questions when human judgement is needed |
| 13 | 5. **Recommend engagement**: Suggest which humans should be involved |
| 14 | 6. **Document uncertainty**: Make missing evidence and conflicts explicit |
| 15 | |
| 16 | ## When to Use |
| 17 | |
| 18 | Use the Planner for any accessibility finding, including: |
| 19 | |
| 20 | * Automated scan results |
| 21 | * Manual testing observations |
| 22 | * Code reviews for accessibility |
| 23 | * Feature design reviews |
| 24 | |
| 25 | The Planner works best early in the accessibility workflow, before implementation or testing occurs. |
| 26 | |
| 27 | ## When Not to Use |
| 28 | |
| 29 | Do not use the Planner for: |
| 30 | |
| 31 | * Direct technical fixes that are clearly required and have no ambiguity |
| 32 | * External accessibility audits that follow different processes |
| 33 | * Quick accessibility compliance checks where no role ownership analysis is needed |
| 34 | |
| 35 | ## Integration with Topic Skills |
| 36 | |
| 37 | The Planner knows which topic skills may be relevant: |
| 38 | |
| 39 | * forms → forms, content-design, keyboard |
| 40 | * keyboard → keyboard, touch-pointer, manual-testing |
| 41 | * color-contrast → color-contrast, light-dark-mode, visual-design |
| 42 | * etc. |
| 43 | |
| 44 | ## ARRM Task Identification |
| 45 | |
| 46 | For a finding, the Planner: |
| 47 | |
| 48 | 1. Maps to relevant WCAG criteria (or ARRM task ID if available) |
| 49 | 2. Identifies Primary, Secondary, and Contributor professional roles |
| 50 | 3. Preserves role ownership assignments from the ARRM data source |
| 51 | 4. Documents any missing or inconsistent role assignments |
| 52 | |
| 53 | ## Decision State Classification |
| 54 | |
| 55 | The Planner uses a criteria-based approach to classify decision states: |
| 56 | |
| 57 | **Deterministic** (auto-fixable): |
| 58 | - Clear normative requirement |
| 59 | - High rule certainty |
| 60 | - Low intent ambiguity |
| 61 | - Low change risk |
| 62 | - High testability |
| 63 | - Meets automation eligibility threshold |
| 64 | |
| 65 | ** Judgement Required**: |
| 66 | - Multiple valid approaches exist |
| 67 | - Trade-offs between accessibility and other requirements |
| 68 | - Need human expertise or user research |
| 69 | - High risk of unintended consequences |
| 70 | |
| 71 | ## Sample Output Structure |
| 72 | |
| 73 | ```json |
| 74 | { |
| 75 | "id": "PLANNER-001", |
| 76 | "title": "Missing alt text on meaningful image", |
| 77 | |
| 78 | "arrm": { |
| 79 | "task_id": "IMG-001", |
| 80 | "wcag_sc": "1.1.1", |
| 81 | "level": "A", |
| 82 | "primary": ["Content Authoring"], |
| 83 | "secondary": ["User Experience Design"], |
| 84 | "contributors": [] |
| 85 | }, |
| 86 | |
| 87 | "decision": { |
| 88 | "state": "cross-role-decision-needed", |
| 89 | "rule_certainty": "high", |
| 90 | "intent_certainty": "medium", |
| 91 | "change_risk": "low", |
| 92 | "testability": "high", |
| 93 | "confidence": "high" |
| 94 | }, |
| 95 | |
| 96 | "ai_assistance": { |
| 97 | "workflows": ["planner", "perspective-auditor"], |
| 98 | "topic_skills": ["image-alt-text"], |
| 99 | "perspectives": ["screen-reader and semantic access"] |
| 100 | }, |
| 101 | |
| 102 | "questions": { |
| 103 | "primary": ["What information does this image add that is not already present in nearby text?"], |
| 104 | "secondary": ["What role does the image play in the user journey?"], |
| 105 | "contributors": [] |
| 106 | }, |
| 107 | |
| 108 | "recommended_engagement": { |
| 109 | "participants": ["Content Authoring", "User Experience Design"], |
| 110 | "reason": "Primary ownership and secondary impact", |
| 111 | "evidence_needed": ["User research on image comprehension", "Business justification for image presence"] |
| 112 | }, |
| 113 | |
| 114 | "escalation": "user-research-needed" |
| 115 | } |
| 116 | ``` |
| 117 | |
| 118 | ## Role-Aware Questions |
| 119 | |
| 120 | The Planner generates questions targeted to specific ARRM roles: |
| 121 | |
| 122 | **Primary Role Questions:** |
| 123 | - Address the core responsibility ownership |
| 124 | - Focus on implementation choices within their domain |
| 125 | - Ask for justification of their preferred approach |
| 126 | |
| 127 | **Secondary Role Questions:** |
| 128 | - Consider broader impact of Primary's decision |
| 129 | - Evaluate interaction with their area of expertise |
| 130 | - Identify potential conflicts with their responsibilities |
| 131 | |
| 132 | **Contributor Role Questions:** |
| 133 | - Ensure the solution works across all relevant technical areas |
| 134 | - Validate implementation feasibility |
| 135 | |
| 136 | ## Conflict Documentation |
| 137 | |
| 138 | When different roles suggest conflicting approaches, the Planner: |
| 139 | |
| 140 | 1. Records each role's concern and recommendation |
| 141 | 2. Documents the evidence supporting each position |
| 142 | 3. Identifies the trade-off being made |
| 143 | 4. Clearly identifies who should make the final decision |
| 144 | 5. Recommends who should be consulted |
| 145 | 6. Notes what evidence would reduce uncertainty |
| 146 | |
| 147 | ## Example Conflict Documentation |
| 148 | |
| 149 | ```json |
| 150 | "role_perspectives": [ |
| 151 | { |
| 152 | "role": "User Experience Design", |
| 153 | "concern": "Increasing spacing may improve precision", |
| 154 | "recommendation": "Add more hit area", |
| 155 | "evidence": "Studies show larger targets reduce errors", |
| 156 | "uncertainty": "Impact on visual grouping not assessed" |
| 157 | }, |
| 158 | { |
| 159 | "role": "Visual Design", |
| 160 | "concern": "Additional spacing may weaken visual grouping", |
| 161 | "recommendation": " |