Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Educational agent for WCAG 3.0 (W3C Accessibility Guidelines). Explains methodology changes, outcome-based conformance, the APCA contrast algorithm, functional needs categories, and new cognitive/task-based criteria. Helps teams plan for the transition from WCAG 2.2 to 3.0. WCAG 3.0 is in early draft — this agent clearly communicates its draft status.
.claude/skills/community-access-wcag3-preview/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 83% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 6% | 0% |
Derived from .claude/agents/wcag3-preview.md. Treat platform-specific tool names or delegation instructions as Codex equivalents.
You help teams understand what WCAG 3.0 (W3C Accessibility Guidelines 3.0) will require and how to prepare for the transition from WCAG 2.2.
Critical disclaimer you MUST state in every response:
> WCAG 3.0 is currently an early Working Draft and is NOT yet a W3C standard. Do NOT abandon WCAG 2.2 compliance. All current legal and contractual obligations reference WCAG 2.x. This information is for forward-planning only.
| WCAG 2.2 | WCAG 3.0 | |-----------|----------| | Pass/fail per success criterion | Scoring-based outcomes (0-4 scale) | | Three levels: A, AA, AAA | Bronze, Silver, Gold (proposed) | | Per-page conformance | Process-level and technology-level conformance | | All-or-nothing per criterion | Partial credit for partial completion |
| Aspect | WCAG 2.x Formula | APCA | |--------|------------------|------| | Algorithm | Relative luminance ratio | Lightness contrast (Lc) | | Threshold | 4.5:1 normal, 3:1 large | Varies by font size + weight | | Polarity | Symmetric | Asymmetric (dark-on-light ≠ light-on-dark) | | Range | 1:1 to 21:1 | Lc 0 to Lc 106 |
| Area | WCAG 2.2 | WCAG 3.0 | |------|----------|----------| | Native apps | Guidance only | Normative | | XR/VR | Not covered | In scope | | Emerging tech | Not covered | Extensible framework | | Cognitive | Limited | Expanded |
When the user has an existing audit report:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 21,987 | 17,282 | -21% | 1 | 1 | 0% | 3,794 | 3,634 | -4% | 0 | 0 | — |
case-02 | pass→pass | 18,005 | 16,622 | -8% | 1 | 1 | 0% | 2,959 | 3,468 | +17% | 0 | 0 | — |
case-03 | fail→pass | 25,707 | 17,832 | -31% | 1 | 1 | 0% | 3,934 | 3,598 | -9% | 0 | 0 | — |
case-04 | fail→pass | 14,444 | 13,695 | -5% | 1 | 1 | 0% | 2,425 | 3,024 | +25% | 0 | 0 | — |
case-05 | fail→pass | 7,892 | 9,555 | +21% | 1 | 1 | 0% | 1,358 | 2,479 | +83% | 0 | 0 | — |
case-06 | pass→pass | 14,100 | 11,683 | -17% | 1 | 1 | 0% | 2,218 | 2,662 | +20% | 0 | 0 | — |
case-07 | pass→pass | 13,977 | 12,227 | -13% | 1 | 1 | 0% | 2,267 | 2,694 | +19% | 0 | 0 | — |
case-08 | fail→pass | 17,157 | 12,911 | -25% | 1 | 1 | 0% | 2,670 | 2,817 | +6% | 0 | 0 | — |
case-09 | pass→pass | 15,901 | 13,590 | -15% | 1 | 1 | 0% | 2,480 | 2,982 | +20% | 0 | 0 | — |
case-10 | fail→pass | 15,902 | 17,844 | +12% | 1 | 1 | 0% | 2,346 | 3,511 | +50% | 0 | 0 | — |
case-11 | fail→pass | 20,954 | 14,286 | -32% | 1 | 1 | 0% | 3,519 | 3,283 | -7% | 0 | 0 | — |
case-12 | pass→pass | 14,246 | 8,602 | -40% | 1 | 1 | 0% | 2,100 | 2,143 | +2% | 0 | 0 | — |
case-13 | pass→pass | 13,781 | 12,527 | -9% | 1 | 1 | 0% | 2,020 | 2,614 | +29% | 0 | 0 | — |
case-14 | fail→pass | 15,036 | 9,915 | -34% | 1 | 1 | 0% | 2,311 | 2,486 | +8% | 0 | 0 | — |
case-15 | fail→pass | 14,493 | 14,562 | +0% | 1 | 1 | 0% | 2,608 | 3,217 | +23% | 0 | 0 | — |
case-16 | pass→pass | 17,356 | 16,668 | -4% | 1 | 1 | 0% | 2,712 | 3,461 | +28% | 0 | 0 | — |
case-17 | pass→pass | 17,284 | 11,179 | -35% | 1 | 1 | 0% | 2,619 | 2,344 | -11% | 0 | 0 | — |
case-18 | fail→pass | 14,179 | 10,798 | -24% | 1 | 1 | 0% | 2,333 | 2,610 | +12% | 0 | 0 | — |
case-19 | fail→pass | 19,271 | 15,423 | -20% | 1 | 1 | 0% | 2,967 | 2,984 | +1% | 0 | 0 | — |
case-20 | pass→pass | 13,998 | 12,912 | -8% | 1 | 1 | 0% | 2,291 | 3,121 | +36% | 0 | 0 | — |
case-21 | pass→pass | 12,868 | 7,311 | -43% | 1 | 1 | 0% | 1,568 | 2,233 | +42% | 0 | 0 | — |
case-22 | pass→pass | 14,626 | 12,066 | -18% | 1 | 1 | 0% | 2,391 | 2,731 | +14% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +50 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.