Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit a live page for accessibility issues, locate each WCAG violation precisely, and return a selector-grounded fix worklist without editing.
.claude/skills/sickn33-accesslint-scan/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-11 | ✓→✗ | ▼ Worse | -53% | 0% |
| case-13 | ✓→✗ | ▼ Worse | -46% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -34% | 0% |
Audit a live page and report what's broken and where. Locate; don't fix. If no URL in $ARGUMENTS, ask for one.
bashPORT=$(npx -y @accesslint/chrome@latest ensure | node -e 'process.stdin.on("data",d=>process.stdout.write(""+JSON.parse(d).port))') npx -y @accesslint/cli@latest "<url>" --port "$PORT" --format json
Flags as needed: --selector, --wait-for "<selector>", --include-aaa, --disable <rules>.
Counts by impact, then one entry per violation:
file:line (symbol) if source is present — never fabricate. If no violation has source, note "source mapping unavailable — located by selector only".NEEDS HUMANDon't edit. For fixes: apply mechanical ones then re-run to verify; for bulk work hand off to accesslint:audit.
bashnpx -y @accesslint/chrome@latest stop --all # skip if ensure reported "managed":false
ensure always determines the port — never hardcode 9222.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-18 | fail→pass | 6,971 | 1,711 | -75% | 1 | 1 | 0% | 1,232 | 729 | -41% | 0 | 0 | — |
case-01 | fail→fail | 9,478 | 7,406 | -22% | 1 | 1 | 0% | 1,692 | 913 | -46% | 0 | 0 | — |
case-02 | fail→fail | 20,985 | 5,360 | -74% | 1 | 1 | 0% | 3,819 | 824 | -78% | 0 | 0 | — |
case-03 | fail→fail | 5,657 | 6,437 | +14% | 1 | 1 | 0% | 321 | 760 | +137% | 0 | 0 | — |
case-04 | fail→fail | 5,262 | 5,683 | +8% | 1 | 1 | 0% | 308 | 1,337 | +334% | 0 | 0 | — |
case-05 | pass→pass | 12,828 | 4,464 | -65% | 1 | 1 | 0% | 2,258 | 1,067 | -53% | 0 | 0 | — |
case-06 | fail→fail | 16,981 | 11,159 | -34% | 1 | 1 | 0% | 3,126 | 2,463 | -21% | 0 | 0 | — |
case-07 | fail→fail | 5,255 | 7,539 | +43% | 1 | 1 | 0% | 893 | 985 | +10% | 0 | 0 | — |
case-08 | fail→fail | 5,759 | 7,019 | +22% | 1 | 1 | 0% | 280 | 1,001 | +258% | 0 | 0 | — |
case-19 | pass→pass | 4,808 | 2,281 | -53% | 1 | 1 | 0% | 725 | 778 | +7% | 0 | 0 | — |
case-09 | pass→pass | 11,903 | 11,133 | -6% | 1 | 1 | 0% | 1,959 | 1,349 | -31% | 0 | 0 | — |
case-10 | fail→fail | 20,922 | 7,135 | -66% | 1 | 1 | 0% | 3,934 | 888 | -77% | 0 | 0 | — |
case-11 | pass→fail | 12,814 | 8,295 | -35% | 1 | 1 | 0% | 2,435 | 1,139 | -53% | 0 | 0 | — |
case-12 | fail→fail | 25,698 | 6,432 | -75% | 1 | 1 | 0% | 4,317 | 772 | -82% | 0 | 0 | — |
case-13 | pass→fail | 7,566 | 6,232 | -18% | 1 | 1 | 0% | 1,535 | 832 | -46% | 0 | 0 | — |
case-14 | pass→pass | 9,283 | 3,480 | -63% | 1 | 1 | 0% | 1,597 | 1,087 | -32% | 0 | 0 | — |
case-15 | pass→pass | 8,775 | 4,691 | -47% | 1 | 1 | 0% | 1,433 | 933 | -35% | 0 | 0 | — |
case-16 | pass→pass | 14,399 | 3,723 | -74% | 1 | 1 | 0% | 2,455 | 1,116 | -55% | 0 | 0 | — |
case-17 | fail→fail | 13,030 | 1,934 | -85% | 1 | 1 | 0% | 2,037 | 677 | -67% | 0 | 0 | — |
case-20 | pass→fail | 13,325 | 6,237 | -53% | 1 | 1 | 0% | 2,326 | 1,543 | -34% | 0 | 0 | — |
case-21 | fail→pass | 17,073 | 6,962 | -59% | 1 | 1 | 0% | 2,218 | 1,545 | -30% | 0 | 0 | — |
case-22 | pass→pass | 13,071 | 6,444 | -51% | 1 | 1 | 0% | 2,169 | 1,673 | -23% | 0 | 0 | — |
case-23 | fail→fail | 18,498 | 10,727 | -42% | 1 | 1 | 0% | 2,910 | 919 | -68% | 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, and 12 counted toward the lift figure. The other 11 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -4 percentage points is the difference between those two pass rates over the 12 comparable cases. 3 cases got worse with the skill loaded, and they are 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.