Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are an accessibility expert specializing in WCAG compliance, inclusive design, and assistive technology compatibility. Conduct audits, identify barriers, and provide remediation guidance.
.claude/skills/sickn33-accessibility-compliance-accessibility-audit/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-09 | ✓→✗ | ▼ Worse | 27% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 16% | 0% |
You are an accessibility expert specializing in WCAG compliance, inclusive design, and assistive technology compatibility. Conduct comprehensive audits, identify barriers, provide remediation guidance, and ensure digital products are accessible to all users.
The user needs to audit and improve accessibility to ensure compliance with WCAG standards and provide an inclusive experience for users with disabilities. Focus on automated testing, manual verification, remediation strategies, and establishing ongoing accessibility practices.
$ARGUMENTS
resources/implementation-playbook.md.resources/implementation-playbook.md for detailed audit steps, tooling, and remediation examples.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | pass→pass | 14,111 | 13,061 | -7% | 1 | 1 | 0% | 2,248 | 2,602 | +16% | 0 | 0 | — |
case-07 | pass→pass | 8,435 | 10,118 | +20% | 1 | 1 | 0% | 1,527 | 2,328 | +52% | 0 | 0 | — |
case-08 | pass→pass | 10,063 | 10,613 | +5% | 1 | 1 | 0% | 1,712 | 2,127 | +24% | 0 | 0 | — |
case-09 | pass→fail | 11,201 | 12,401 | +11% | 1 | 1 | 0% | 2,023 | 2,560 | +27% | 0 | 0 | — |
case-01 | fail→fail | 43,203 | 32,922 | -24% | 1 | 1 | 0% | 8,228 | 6,240 | -24% | 0 | 0 | — |
case-02 | fail→fail | 40,372 | 40,504 | +0% | 1 | 1 | 0% | 8,242 | 8,612 | +4% | 0 | 0 | — |
case-03 | fail→pass | 10,468 | 3,907 | -63% | 1 | 1 | 0% | 1,766 | 1,001 | -43% | 0 | 0 | — |
case-04 | fail→pass | 17,045 | 4,985 | -71% | 1 | 1 | 0% | 2,817 | 1,242 | -56% | 0 | 0 | — |
case-05 | fail→fail | 15,691 | 17,931 | +14% | 1 | 1 | 0% | 3,003 | 3,677 | +22% | 0 | 0 | — |
case-10 | pass→pass | 9,379 | 11,061 | +18% | 1 | 1 | 0% | 1,809 | 2,449 | +35% | 0 | 0 | — |
case-11 | pass→pass | 11,768 | 10,293 | -13% | 1 | 1 | 0% | 1,931 | 2,050 | +6% | 0 | 0 | — |
case-12 | pass→pass | 13,007 | 6,682 | -49% | 1 | 1 | 0% | 2,200 | 1,478 | -33% | 0 | 0 | — |
case-13 | pass→pass | 12,278 | 16,117 | +31% | 1 | 1 | 0% | 2,061 | 3,326 | +61% | 0 | 0 | — |
case-14 | pass→pass | 9,479 | 11,602 | +22% | 1 | 1 | 0% | 1,774 | 2,390 | +35% | 0 | 0 | — |
case-15 | pass→pass | 14,887 | 19,578 | +32% | 1 | 1 | 0% | 2,476 | 3,854 | +56% | 0 | 0 | — |
case-16 | pass→pass | 11,130 | 14,768 | +33% | 1 | 1 | 0% | 1,891 | 2,963 | +57% | 0 | 0 | — |
case-17 | pass→pass | 7,107 | 9,798 | +38% | 1 | 1 | 0% | 1,210 | 1,994 | +65% | 0 | 0 | — |
case-18 | pass→pass | 12,159 | 15,828 | +30% | 1 | 1 | 0% | 2,221 | 3,173 | +43% | 0 | 0 | — |
case-19 | fail→pass | 15,612 | 4,874 | -69% | 1 | 1 | 0% | 2,706 | 1,198 | -56% | 0 | 0 | — |
case-20 | pass→pass | 10,761 | 11,026 | +2% | 1 | 1 | 0% | 2,026 | 2,361 | +17% | 0 | 0 | — |
case-21 | pass→pass | 11,901 | 8,600 | -28% | 1 | 1 | 0% | 2,064 | 1,939 | -6% | 0 | 0 | — |
case-22 | pass→pass | 12,281 | 11,495 | -6% | 1 | 1 | 0% | 1,941 | 2,631 | +36% | 0 | 0 | — |
case-23 | pass→pass | 5,354 | 6,425 | +20% | 1 | 1 | 0% | 921 | 1,463 | +59% | 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. 23 cases were attempted. The headline lift of +9 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.