Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit a Cursor plugin for marketplace readiness. Use when validating manifests, component metadata, discovery paths, and submission quality before publishing.
.claude/skills/kunanonj-cursor-plugin-create-review-plugin-submission/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -23% | 0% |
A plugin is implemented and needs a final quality check before submission or release.
.cursor-plugin/plugin.json existsname is valid lowercase kebab-casedescription, version, author, license)skills/*/SKILL.mdrules/ as .mdc or markdown variantsagents/ markdown filescommands/ markdown or text fileshooks/hooks.jsonmcp.json (or mcpServers override)name and description frontmattername and description.cursor-plugin/marketplace.jsonsource resolves to plugin directory and names are uniqueREADME.md states purpose, installation, and component coverage| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | fail→pass | 11,864 | 5,372 | -55% | 1 | 1 | 0% | 2,348 | 1,491 | -36% | 0 | 0 | — |
case-01 | fail→fail | 14,558 | 1,414 | -90% | 1 | 1 | 0% | 2,436 | 657 | -73% | 0 | 0 | — |
case-02 | fail→fail | 10,658 | 1,590 | -85% | 1 | 1 | 0% | 2,180 | 666 | -69% | 0 | 0 | — |
case-03 | fail→fail | 15,191 | 1,548 | -90% | 1 | 1 | 0% | 2,752 | 636 | -77% | 0 | 0 | — |
case-04 | pass→pass | 8,879 | 4,355 | -51% | 1 | 1 | 0% | 1,739 | 1,332 | -23% | 0 | 0 | — |
case-05 | pass→pass | 9,063 | 6,404 | -29% | 1 | 1 | 0% | 2,076 | 1,685 | -19% | 0 | 0 | — |
case-07 | fail→pass | 9,923 | 3,506 | -65% | 1 | 1 | 0% | 1,801 | 1,061 | -41% | 0 | 0 | — |
case-08 | fail→pass | 8,372 | 2,805 | -66% | 1 | 1 | 0% | 1,534 | 980 | -36% | 0 | 0 | — |
case-09 | pass→pass | 10,095 | 4,917 | -51% | 1 | 1 | 0% | 1,719 | 1,350 | -21% | 0 | 0 | — |
case-10 | pass→pass | 8,019 | 3,328 | -58% | 1 | 1 | 0% | 1,507 | 939 | -38% | 0 | 0 | — |
case-11 | pass→pass | 3,876 | 2,459 | -37% | 1 | 1 | 0% | 776 | 781 | +1% | 0 | 0 | — |
case-12 | fail→pass | 7,075 | 2,823 | -60% | 1 | 1 | 0% | 1,263 | 923 | -27% | 0 | 0 | — |
case-13 | pass→pass | 9,322 | 3,184 | -66% | 1 | 1 | 0% | 1,939 | 1,108 | -43% | 0 | 0 | — |
case-14 | pass→pass | 7,234 | 5,505 | -24% | 1 | 1 | 0% | 1,411 | 1,481 | +5% | 0 | 0 | — |
case-15 | pass→pass | 9,455 | 5,847 | -38% | 1 | 1 | 0% | 1,726 | 1,488 | -14% | 0 | 0 | — |
case-16 | pass→pass | 6,489 | 3,124 | -52% | 1 | 1 | 0% | 1,127 | 1,012 | -10% | 0 | 0 | — |
case-17 | pass→pass | 20,447 | 6,262 | -69% | 1 | 1 | 0% | 2,099 | 1,567 | -25% | 0 | 0 | — |
case-18 | pass→pass | 4,929 | 3,691 | -25% | 1 | 1 | 0% | 973 | 972 | -0% | 0 | 0 | — |
case-19 | pass→pass | 7,631 | 4,860 | -36% | 1 | 1 | 0% | 1,449 | 1,155 | -20% | 0 | 0 | — |
case-20 | pass→pass | 12,494 | 6,581 | -47% | 1 | 1 | 0% | 2,248 | 1,652 | -27% | 0 | 0 | — |
case-21 | pass→pass | 16,157 | 15,315 | -5% | 1 | 1 | 0% | 3,896 | 3,548 | -9% | 0 | 0 | — |
case-22 | pass→pass | 15,176 | 14,025 | -8% | 1 | 1 | 0% | 3,547 | 3,601 | +2% | 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 +18 percentage points is the difference between those two pass rates over the 22 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.