Planner Workflow
The Planner identifies and categorizes accessibility findings, making them actionable for human professionals while maximizing automation where possible.
Purpose
The Planner's role is to:
- Map findings to ARRM tasks: Identify which ARRM professional roles own responsibility for the finding
- Distinguish automation: Separate clear-rule issues from judgement calls
- Identify perspectives: Consider impacts on different user groups
- Generate questions: Ask role-specific questions when human judgement is needed
- Recommend engagement: Suggest which humans should be involved
- Document uncertainty: Make missing evidence and conflicts explicit
When to Use
Use the Planner for any accessibility finding, including:
- Automated scan results
- Manual testing observations
- Code reviews for accessibility
- Feature design reviews
The Planner works best early in the accessibility workflow, before implementation or testing occurs.
When Not to Use
Do not use the Planner for:
- Direct technical fixes that are clearly required and have no ambiguity
- External accessibility audits that follow different processes
- Quick accessibility compliance checks where no role ownership analysis is needed
Integration with Topic Skills
The Planner knows which topic skills may be relevant:
- forms → forms, content-design, keyboard
- keyboard → keyboard, touch-pointer, manual-testing
- color-contrast → color-contrast, light-dark-mode, visual-design
- etc.
ARRM Task Identification
For a finding, the Planner:
- Maps to relevant WCAG criteria (or ARRM task ID if available)
- Identifies Primary, Secondary, and Contributor professional roles
- Preserves role ownership assignments from the ARRM data source
- Documents any missing or inconsistent role assignments
Decision State Classification
The Planner uses a criteria-based approach to classify decision states:
Deterministic (auto-fixable):
- Clear normative requirement
- High rule certainty
- Low intent ambiguity
- Low change risk
- High testability
- Meets automation eligibility threshold
** Judgement Required**:
- Multiple valid approaches exist
- Trade-offs between accessibility and other requirements
- Need human expertise or user research
- High risk of unintended consequences
Sample Output Structure
{
"id": "PLANNER-001",
"title": "Missing alt text on meaningful image",
"arrm": {
"task_id": "IMG-001",
"wcag_sc": "1.1.1",
"level": "A",
"primary": ["Content Authoring"],
"secondary": ["User Experience Design"],
"contributors": []
},
"decision": {
"state": "cross-role-decision-needed",
"rule_certainty": "high",
"intent_certainty": "medium",
"change_risk": "low",
"testability": "high",
"confidence": "high"
},
"ai_assistance": {
"workflows": ["planner", "perspective-auditor"],
"topic_skills": ["image-alt-text"],
"perspectives": ["screen-reader and semantic access"]
},
"questions": {
"primary": ["What information does this image add that is not already present in nearby text?"],
"secondary": ["What role does the image play in the user journey?"],
"contributors": []
},
"recommended_engagement": {
"participants": ["Content Authoring", "User Experience Design"],
"reason": "Primary ownership and secondary impact",
"evidence_needed": ["User research on image comprehension", "Business justification for image presence"]
},
"escalation": "user-research-needed"
}
Role-Aware Questions
The Planner generates questions targeted to specific ARRM roles:
Primary Role Questions:
- Address the core responsibility ownership
- Focus on implementation choices within their domain
- Ask for justification of their preferred approach
Secondary Role Questions:
- Consider broader impact of Primary's decision
- Evaluate interaction with their area of expertise
- Identify potential conflicts with their responsibilities
Contributor Role Questions:
- Ensure the solution works across all relevant technical areas
- Validate implementation feasibility
Conflict Documentation
When different roles suggest conflicting approaches, the Planner:
- Records each role's concern and recommendation
- Documents the evidence supporting each position
- Identifies the trade-off being made
- Clearly identifies who should make the final decision
- Recommends who should be consulted
- Notes what evidence would reduce uncertainty
Example Conflict Documentation
"role_perspectives": [
{
"role": "User Experience Design",
"concern": "Increasing spacing may improve precision",
"recommendation": "Add more hit area",
"evidence": "Studies show larger targets reduce errors",
"uncertainty": "Impact on visual grouping not assessed"
},
{
"role": "Visual Design",
"concern": "Additional spacing may weaken visual grouping",
"recommendation": "Maintain current spacing",
"evidence": "Visual hierarchy depends on proximity",
"uncertainty": "Trade-off between accessibility and aesthetics"
}
],
"next_step": "Front-End Development should prototype the implementation",
"consultation": ["Visual Design should review affordance", "UX should validate task clarity"]
Integration with Other Workflows
The Planner works with:
- Critic: The Planner can inform the Critic's review of decisions
- Tester: The Planner identifies which tests should be run for each finding
- Perspective Auditor: The Planner identifies which user perspectives to examine
- Bug Reporter: The Planner ensures Bug Reports preserve ARRM ownership
Decision State Escalation Paths
| State | Recommended Action |
|---|---|
| auto-fixable | Apply fix, run tests, document |
| human-confirmation-needed | Human validation, record response |
| cross-role-decision-needed | Facilitate cross-role discussion |
| user-research-needed | Arrange disabled-user testing |
| specialist-review-needed | Consult relevant external expert |
| governance-decision-needed | Escalate to policy/legal review |
| insufficient-evidence | Document missing evidence, ask targeted questions |
The Planner ensures that ambiguous or complex accessibility work is routed appropriately, with clear visibility into who owns what decisions and why.