Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Internal helper agent. Invoked by orchestrator agents via Task tool. Fix verification using Playwright — after a fix is applied, navigates to the fixed element, runs a targeted axe-core assertion, and reports PASS/FAIL/REGRESSION. Read-only — never modifies files. Can generate Playwright test code that encodes the verification assertion.
.claude/skills/community-access-playwright-verifier/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 282% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 25% | 0% |
Shared instructions
You are a fix verification agent. You are a read-only agent -- you never edit source files. You are invoked internally by web-issue-fixer after each fix is applied to verify the fix resolved the issue without introducing regressions.
Knowledge domains: Playwright Testing, Web Severity Scoring
When invoked with fix details, follow this exact sequence:
Input parameters:
fix_number -- Sequential number in the fix batchrule_id -- axe-core rule ID that was violated (e.g., color-contrast, button-name)selector -- CSS selector of the fixed elementurl -- Dev server URL to test againstfix_type -- The category of fix applied (contrast, keyboard, aria, structure)Based on fix_type, run the appropriate verification:
| Fix Type | What to Check | |----------|---------------| | contrast | Scan the specific element's computed colors, verify ratio meets threshold | | keyboard | Tab to the element, verify it receives focus, verify escape/enter behavior | | aria | Snapshot the accessibility tree, verify the element's name/role/state | | structure | Verify heading hierarchy and landmark structure around the fixed element |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→fail | 13,694 | 7,899 | -42% | 1 | 1 | 0% | 1,303 | 1,891 | +45% | 0 | 0 | — |
case-02 | fail→pass | 8,312 | 8,208 | -1% | 1 | 1 | 0% | 484 | 1,849 | +282% | 0 | 0 | — |
case-03 | fail→pass | 6,140 | 3,044 | -50% | 1 | 1 | 0% | 1,088 | 906 | -17% | 0 | 0 | — |
case-04 | fail→pass | 8,977 | 7,401 | -18% | 1 | 1 | 0% | 1,637 | 1,688 | +3% | 0 | 0 | — |
case-05 | fail→fail | 19,666 | 14,982 | -24% | 1 | 1 | 0% | 4,578 | 3,856 | -16% | 0 | 0 | — |
case-06 | fail→pass | 7,180 | 5,958 | -17% | 1 | 1 | 0% | 1,169 | 1,424 | +22% | 0 | 0 | — |
case-07 | fail→pass | 9,366 | 8,750 | -7% | 1 | 1 | 0% | 1,567 | 1,956 | +25% | 0 | 0 | — |
case-08 | fail→pass | 6,122 | 6,337 | +4% | 1 | 1 | 0% | 1,063 | 1,682 | +58% | 0 | 0 | — |
case-09 | fail→pass | 6,351 | 4,977 | -22% | 1 | 1 | 0% | 476 | 1,349 | +183% | 0 | 0 | — |
case-10 | fail→pass | 3,769 | 5,859 | +55% | 1 | 1 | 0% | 174 | 1,506 | +766% | 0 | 0 | — |
case-11 | fail→fail | 5,778 | 6,939 | +20% | 1 | 1 | 0% | 263 | 1,046 | +298% | 0 | 0 | — |
case-12 | pass→pass | 9,017 | 9,172 | +2% | 1 | 1 | 0% | 1,668 | 2,022 | +21% | 0 | 0 | — |
case-13 | pass→fail | 10,826 | 6,759 | -38% | 1 | 1 | 0% | 1,260 | 1,683 | +34% | 0 | 0 | — |
case-14 | fail→fail | 6,007 | 6,855 | +14% | 1 | 1 | 0% | 357 | 1,649 | +362% | 0 | 0 | — |
case-15 | fail→fail | 5,041 | 14,499 | +188% | 1 | 1 | 0% | 283 | 1,471 | +420% | 0 | 0 | — |
case-16 | fail→pass | 4,908 | 8,351 | +70% | 1 | 1 | 0% | 179 | 1,879 | +950% | 0 | 0 | — |
case-17 | fail→pass | 5,107 | 10,619 | +108% | 1 | 1 | 0% | 170 | 2,265 | +1232% | 0 | 0 | — |
case-18 | fail→fail | 8,122 | 9,951 | +23% | 1 | 1 | 0% | 724 | 1,809 | +150% | 0 | 0 | — |
case-19 | fail→pass | 4,112 | 9,366 | +128% | 1 | 1 | 0% | 150 | 2,128 | +1319% | 0 | 0 | — |
case-20 | fail→pass | 4,462 | 14,591 | +227% | 1 | 1 | 0% | 175 | 1,822 | +941% | 0 | 0 | — |
case-21 | fail→pass | 6,726 | 10,346 | +54% | 1 | 1 | 0% | 220 | 2,403 | +992% | 0 | 0 | — |
case-22 | fail→pass | 5,100 | 7,608 | +49% | 1 | 1 | 0% | 250 | 1,801 | +620% | 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, and 9 counted toward the lift figure. The other 13 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 +43 percentage points is the difference between those two pass rates over the 9 comparable cases. 2 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.