Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →小红书美妆内容创作技能，支持笔记生成、笔记优化、文生图描述创作；适用于美妆护肤产品的内容创作、测评和营销推广
.claude/skills/anbeime-xiaohongshu-makeup/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 282% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 76% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 92% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 117% | 0% |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | fail→pass | 4,196 | 5,751 | +37% | 1 | 1 | 0% | 779 | 2,978 | +282% | 0 | 0 | — |
case-01 | fail→pass | 12,681 | 10,685 | -16% | 1 | 1 | 0% | 2,160 | 3,794 | +76% | 0 | 0 | — |
case-02 | fail→pass | 14,993 | 16,172 | +8% | 1 | 1 | 0% | 2,430 | 4,655 | +92% | 0 | 0 | — |
case-03 | pass→pass | 11,006 | 10,696 | -3% | 1 | 1 | 0% | 2,041 | 3,833 | +88% | 0 | 0 | — |
case-08 | fail→pass | 10,412 | 12,092 | +16% | 1 | 1 | 0% | 1,812 | 3,797 | +110% | 0 | 0 | — |
case-04 | pass→pass | 13,352 | 8,170 | -39% | 1 | 1 | 0% | 2,244 | 3,311 | +48% | 0 | 0 | — |
case-05 | fail→fail | 13,247 | 11,147 | -16% | 1 | 1 | 0% | 2,259 | 3,808 | +69% | 0 | 0 | — |
case-06 | fail→fail | 20,526 | 21,739 | +6% | 1 | 1 | 0% | 3,327 | 5,600 | +68% | 0 | 0 | — |
case-07 | pass→pass | 12,291 | 13,030 | +6% | 1 | 1 | 0% | 1,907 | 4,170 | +119% | 0 | 0 | — |
case-09 | pass→pass | 8,575 | 7,402 | -14% | 1 | 1 | 0% | 1,408 | 3,168 | +125% | 0 | 0 | — |
case-10 | pass→pass | 11,481 | 14,343 | +25% | 1 | 1 | 0% | 1,862 | 4,212 | +126% | 0 | 0 | — |
case-12 | pass→pass | 14,413 | 8,895 | -38% | 1 | 1 | 0% | 2,414 | 3,440 | +43% | 0 | 0 | — |
case-13 | fail→pass | 10,972 | 11,479 | +5% | 1 | 1 | 0% | 1,847 | 4,003 | +117% | 0 | 0 | — |
case-14 | pass→pass | 14,891 | 12,173 | -18% | 1 | 1 | 0% | 2,522 | 3,872 | +54% | 0 | 0 | — |
case-15 | pass→pass | 8,498 | 9,654 | +14% | 1 | 1 | 0% | 1,489 | 3,539 | +138% | 0 | 0 | — |
case-16 | pass→pass | 10,873 | 9,451 | -13% | 1 | 1 | 0% | 1,897 | 3,485 | +84% | 0 | 0 | — |
case-17 | pass→pass | 15,047 | 15,762 | +5% | 1 | 1 | 0% | 2,535 | 4,534 | +79% | 0 | 0 | — |
case-18 | fail→pass | 14,758 | 14,583 | -1% | 1 | 1 | 0% | 2,386 | 4,416 | +85% | 0 | 0 | — |
case-19 | pass→pass | 11,398 | 7,696 | -32% | 1 | 1 | 0% | 1,810 | 3,063 | +69% | 0 | 0 | — |
case-20 | pass→pass | 18,048 | 17,581 | -3% | 1 | 1 | 0% | 2,756 | 4,808 | +74% | 0 | 0 | — |
case-21 | fail→pass | 14,835 | 12,134 | -18% | 1 | 1 | 0% | 2,145 | 3,796 | +77% | 0 | 0 | — |
case-22 | pass→pass | 14,726 | 9,949 | -32% | 1 | 1 | 0% | 2,583 | 3,540 | +37% | 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 +32 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.