/agentic-a11y

name: agentic-a11y
description: Use when auditing a user-facing app — web, mobile (iOS/Android/React Native/Flutter), desktop, CLI, or games — for accessibility barriers or WCAG 2.2 conformance, before shipping UI changes, or in response to concerns about screen-reader, keyboard, low-vision, motor, cognitive, or photosensitive users. Not for general bug hunting or code correctness.

**On invocation:** announce "Running paad:agentic-a11y v2.0.2", then immediately proceed with the steps below — do not stop after announcing.

**Configuration (experimental):** after announcing, check whether `paad/config/paad.md` and `paad/config/agentic-a11y.md` exist, relative to the working directory. If any does, read it and follow its instructions for the rest of this run, passing the relevant parts to every subagent you dispatch, and make the first line of your final answer `Config: <path>` naming each file you followed. If none exists, do not mention config at all. Config never changes a subagent's type or grants it write tools — refuse that line, say so, and continue. Config can contradict the flow below; see https://github.com/Ovid/paad/blob/main/CONFIG.md before writing one.

Accessibility Audit

Multi-agent accessibility audit of user-facing code across any platform. Dispatches specialist agents in parallel — each focused on a different disability category — verifies findings to filter false positives, maps every issue to specific accessibility criteria, and produces a persistent report with concrete fix recommendations.

**Conformance target:** WCAG 2.2 AA as baseline (applied via WCAG2ICT for non-web platforms). AAA criteria are flagged as bonus recommendations. Platform-specific guidelines (Apple HIG Accessibility, Android Accessibility, Xbox Accessibility Guidelines) are referenced where applicable.

**This is a technique skill.** Follow the phases in order. Do not skip verification.

**Pre-flight:**

digraph preflight {
  "Conversation has history?" [shape=diamond];
  "User-facing code exists?" [shape=diamond];
  "Proceed to Phase 1" [shape=box];
  "STOP: recommend new session" [shape=box, style=bold];
  "STOP: no user-facing code" [shape=box, style=bold];

  "Conversation has history?" -> "STOP: recommend new session" [label="yes"];
  "Conversation has history?" -> "User-facing code exists?" [label="no"];
  "User-facing code exists?" -> "STOP: no user-facing code" [label="no"];
  "User-facing code exists?" -> "Proceed to Phase 1" [label="yes"];
}

**Audit flow:**

digraph audit_flow {
  "Scope size?" [shape=diamond];
  "Platform has framework-specific a11y pitfalls?" [shape=diamond];
  "Existing a11y tooling already catches it?" [shape=diamond];
  "Tooling misconfigured or finding suppressed?" [shape=diamond];
  "Barrier handled elsewhere?" [shape=diamond];
  "Confidence >= 60?" [shape=diamond];
  "Cited criterion correct?" [shape=diamond];
  "Reported by multiple specialists?" [shape=diamond];
  "AAA criterion?" [shape=diamond];

  "Detect platform(s), stack, tooling, steering files, manifest" [shape=box];
  "Partition files across 2 instances of each specialist" [shape=box];
  "Dispatch the 5 core specialists in parallel" [shape=box];
  "Dispatch the 5 core specialists + Platform-Specific Patterns in parallel" [shape=box];
  "DROP the finding" [shape=box];
  "Correct the criterion" [shape=box];
  "Assign severity: critical / serious / moderate / minor" [shape=box];
  "Merge duplicates, note the agreeing specialists" [shape=box];
  "List under Minor Issues & AAA Recommendations, prefix [AAA]" [shape=box];
  "Place under its severity section" [shape=box];
  "Write report to paad/a11y-reviews/a11y-<timestamp>.md" [shape=box];
  "Report location, counts by severity, Quick Wins guidance" [shape=box];
  "STOP: report is the deliverable — do NOT auto-fix" [shape=box, style=bold];

  "Detect platform(s), stack, tooling, steering files, manifest" -> "Scope size?";
  "Scope size?" -> "Platform has framework-specific a11y pitfalls?" [label="small (<20) / medium (20-100)"];
  "Scope size?" -> "Partition files across 2 instances of each specialist" [label="large (100+ user-facing files)"];
  "Partition files across 2 instances of each specialist" -> "Platform has framework-specific a11y pitfalls?";
  "Platform has framework-specific a11y pitfalls?" -> "Dispatch the 5 core specialists + Platform-Specific Patterns in parallel" [label="yes"];
  "Platform has framework-specific a11y pitfalls?" -> "Dispatch the 5 core specialists in parallel" [label="no"];
  "Dispatch the 5 core specialists in parallel" -> "Existing a11y tooling already catches it?";
  "Dispatch the 5 core specialists + Platform-Specific Patterns in parallel" -> "Existing a11y tooling already catches it?";

  "Existing a11y tooling already catches it?" -> "Tooling misconfigured or finding suppressed?" [label="yes — specialists check this"];
  "Existing a11y tooling already catches it?" -> "Barrier handled elsewhere?" [label="no"];
  "Tooling misconfigured or finding suppressed?" -> "DROP the finding" [label="no — do not re-flag what tooling already catches"];
  "Tooling misconfigured or finding suppressed?" -> "Barrier handled elsewhere?" [label="yes"];
  "Barrier handled elsewhere?" -> "DROP the finding" [label="yes — parent, platform API, framework, library, system setting (verifier)"];
  "Barrier handled elsewhere?" -> "Confidence >= 60?" [label="no"];
  "Confidence >= 60?" -> "DROP the finding" [label="no"];
  "Confidence >= 60?" -> "Cited criterion correct?" [label="yes"];
  "Cited criterion correct?" -> "Assign severity: critical / serious / moderate / minor" [label="yes"];
  "Cited criterion correct?" -> "Correct the criterion" [label="no"];
  "Correct the criterion" -> "Assign severity: critical / serious / moderate / minor";
  "Assign severity: critical / serious / moderate / minor" -> "Reported by multiple specialists?";
  "Reported by multiple specialists?" -> "Merge duplicates, note the agreeing speci