Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →电影感多平台社媒分析: 互动图、悬浮洞察、区间对比、明暗主题
.claude/skills/nexu-io-social-media-matrix/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-03 | ✓→✗ | ▼ Worse | 50% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 44% | 0% |
| case-07 | ✓→✗ | ▼ Worse | 146% | 0% |
| case-10 | ✓→✗ | ▼ Worse | 21% | 0% |
【模板: 社媒矩阵追踪 (Social Matrix Tracker)】 【意图】电影感、数据密集的多平台社媒看板。 【布局】
【设计细节】
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 26,762 | 27,436 | +3% | 1 | 1 | 0% | 6,221 | 6,334 | +2% | 0 | 0 | — |
case-02 | fail→fail | 31,555 | 30,747 | -3% | 1 | 1 | 0% | 6,226 | 6,339 | +2% | 0 | 0 | — |
case-03 | pass→fail | 20,936 | 27,120 | +30% | 1 | 1 | 0% | 4,228 | 6,326 | +50% | 0 | 0 | — |
case-04 | pass→fail | 23,202 | 27,853 | +20% | 1 | 1 | 0% | 4,393 | 6,336 | +44% | 0 | 0 | — |
case-05 | pass→pass | 20,109 | 29,275 | +46% | 1 | 1 | 0% | 3,323 | 6,333 | +91% | 0 | 0 | — |
case-06 | pass→pass | 23,446 | 23,507 | +0% | 1 | 1 | 0% | 4,080 | 4,165 | +2% | 0 | 0 | — |
case-07 | pass→fail | 15,088 | 29,156 | +93% | 1 | 1 | 0% | 2,561 | 6,311 | +146% | 0 | 0 | — |
case-08 | fail→pass | 23,158 | 22,375 | -3% | 1 | 1 | 0% | 4,663 | 4,870 | +4% | 0 | 0 | — |
case-09 | pass→pass | 16,024 | 27,074 | +69% | 1 | 1 | 0% | 2,700 | 6,306 | +134% | 0 | 0 | — |
case-10 | pass→fail | 24,395 | 27,765 | +14% | 1 | 1 | 0% | 5,228 | 6,301 | +21% | 0 | 0 | — |
case-11 | pass→pass | 29,936 | 29,294 | -2% | 1 | 1 | 0% | 6,065 | 6,306 | +4% | 0 | 0 | — |
case-12 | pass→fail | 28,702 | 31,068 | +8% | 1 | 1 | 0% | 6,194 | 6,307 | +2% | 0 | 0 | — |
case-13 | fail→fail | 30,467 | 29,727 | -2% | 1 | 1 | 0% | 6,193 | 6,306 | +2% | 0 | 0 | — |
case-14 | pass→pass | 19,371 | 26,937 | +39% | 1 | 1 | 0% | 4,020 | 5,996 | +49% | 0 | 0 | — |
case-15 | pass→pass | 18,424 | 28,247 | +53% | 1 | 1 | 0% | 3,844 | 6,306 | +64% | 0 | 0 | — |
case-16 | pass→pass | 23,465 | 27,330 | +16% | 1 | 1 | 0% | 5,372 | 6,310 | +17% | 0 | 0 | — |
case-17 | pass→pass | 25,619 | 29,838 | +16% | 1 | 1 | 0% | 3,885 | 6,301 | +62% | 0 | 0 | — |
case-18 | pass→fail | 24,801 | 30,266 | +22% | 1 | 1 | 0% | 4,944 | 6,303 | +27% | 0 | 0 | — |
case-19 | pass→pass | 23,076 | 28,532 | +24% | 1 | 1 | 0% | 4,694 | 6,299 | +34% | 0 | 0 | — |
case-20 | fail→fail | 29,216 | 29,953 | +3% | 1 | 1 | 0% | 6,181 | 6,294 | +2% | 0 | 0 | — |
case-21 | pass→pass | 23,596 | 26,501 | +12% | 1 | 1 | 0% | 4,949 | 6,192 | +25% | 0 | 0 | — |
case-22 | pass→fail | 20,801 | 25,633 | +23% | 1 | 1 | 0% | 4,087 | 6,316 | +55% | 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 -40 percentage points is the difference between those two pass rates over the 22 comparable cases. 7 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.