Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Safely searches and installs agent skills from public marketplaces (e.g.agen tskill.sh) with user consent and security checks. Use when user says findski ll, install skill from marketplace, /learn. Triggers: find skill, installski ll, skill marketplace, check skill safety.
.claude/skills/stijnman-skill-marketplace-installer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -67% | 0% |
skill-rubric-reviewerhitl-approverprivacy-redactor| Failure | Response | |---------|----------| | CLI unavailable | Document manual install from source URL with rubric review first. | | Scan flags risk | Block install; explain finding; offer alternatives. |
This skill is designed for public distribution. Constraints:
Input: User request matching triggers above. Output: Structured result per workflow; local artifacts only unless user opts in.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→pass | 14,970 | 9,528 | -36% | 1 | 1 | 0% | 1,976 | 2,082 | +5% | 0 | 0 | — |
case-01 | fail→fail | 3,090 | 15,921 | +415% | 1 | 1 | 0% | 405 | 1,904 | +370% | 0 | 0 | — |
case-02 | fail→pass | 15,110 | 8,164 | -46% | 1 | 1 | 0% | 1,435 | 1,775 | +24% | 0 | 0 | — |
case-03 | fail→pass | 11,291 | 7,050 | -38% | 1 | 1 | 0% | 1,542 | 1,706 | +11% | 0 | 0 | — |
case-04 | pass→fail | 10,970 | 10,586 | -4% | 1 | 1 | 0% | 1,988 | 2,377 | +20% | 0 | 0 | — |
case-05 | pass→pass | 13,608 | 10,194 | -25% | 1 | 1 | 0% | 2,351 | 2,199 | -6% | 0 | 0 | — |
case-06 | pass→pass | 13,297 | 11,257 | -15% | 1 | 1 | 0% | 2,399 | 2,405 | +0% | 0 | 0 | — |
case-07 | pass→pass | 7,239 | 5,058 | -30% | 1 | 1 | 0% | 983 | 1,308 | +33% | 0 | 0 | — |
case-09 | pass→pass | 9,537 | 11,955 | +25% | 1 | 1 | 0% | 1,249 | 1,425 | +14% | 0 | 0 | — |
case-10 | pass→pass | 9,234 | 5,006 | -46% | 1 | 1 | 0% | 1,467 | 1,171 | -20% | 0 | 0 | — |
case-11 | pass→pass | 5,225 | 7,564 | +45% | 1 | 1 | 0% | 630 | 1,701 | +170% | 0 | 0 | — |
case-12 | fail→pass | 9,848 | 2,331 | -76% | 1 | 1 | 0% | 1,328 | 774 | -42% | 0 | 0 | — |
case-13 | fail→fail | 13,285 | 9,624 | -28% | 1 | 1 | 0% | 1,990 | 1,598 | -20% | 0 | 0 | — |
case-14 | fail→fail | 6,781 | 2,032 | -70% | 1 | 1 | 0% | 1,079 | 795 | -26% | 0 | 0 | — |
case-15 | fail→pass | 12,659 | 3,157 | -75% | 1 | 1 | 0% | 2,210 | 731 | -67% | 0 | 0 | — |
case-16 | fail→pass | 11,952 | 2,657 | -78% | 1 | 1 | 0% | 1,840 | 822 | -55% | 0 | 0 | — |
case-17 | pass→pass | 6,378 | 5,104 | -20% | 1 | 1 | 0% | 1,016 | 1,102 | +8% | 0 | 0 | — |
case-18 | pass→pass | 9,018 | 4,326 | -52% | 1 | 1 | 0% | 1,398 | 1,182 | -15% | 0 | 0 | — |
case-19 | pass→fail | 9,573 | 2,489 | -74% | 1 | 1 | 0% | 1,618 | 839 | -48% | 0 | 0 | — |
case-20 | pass→pass | 10,769 | 2,458 | -77% | 1 | 1 | 0% | 1,662 | 795 | -52% | 0 | 0 | — |
case-21 | fail→pass | 16,966 | 2,116 | -88% | 1 | 1 | 0% | 2,224 | 813 | -63% | 0 | 0 | — |
case-22 | fail→pass | 16,527 | 2,409 | -85% | 1 | 1 | 0% | 2,142 | 809 | -62% | 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 +20 percentage points is the difference between those two pass rates over the 22 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.