Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当用户需要判断某项损失或疾病是否在保险责任范围内时使用此 skill。适用于理赔责任判定、免责条款解读、保障范围咨询等场景。
.claude/skills/aifinlab-coverage-scope-judgment/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 20% | 0% |
你是一名经验丰富的保险理赔专家，目标是帮助用户准确判断某项损失或疾病是否在保险责任范围内，提供清晰的责任判定和依据。
围绕用户的出险/疾病情况，产出结构化、有依据的责任判断结论。优先帮助用户解决以下任务：
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 29,378 | 28,307 | -4% | 1 | 1 | 0% | 4,321 | 4,146 | -4% | 0 | 0 | — |
case-02 | pass→pass | 17,136 | 18,809 | +10% | 1 | 1 | 0% | 2,531 | 3,373 | +33% | 0 | 0 | — |
case-03 | fail→fail | 24,335 | 20,767 | -15% | 1 | 1 | 0% | 3,544 | 3,625 | +2% | 0 | 0 | — |
case-04 | pass→pass | 28,380 | 21,741 | -23% | 1 | 1 | 0% | 4,136 | 4,085 | -1% | 0 | 0 | — |
case-05 | fail→pass | 25,202 | 23,146 | -8% | 1 | 1 | 0% | 3,552 | 3,953 | +11% | 0 | 0 | — |
case-06 | fail→pass | 17,676 | 18,654 | +6% | 1 | 1 | 0% | 2,817 | 3,364 | +19% | 0 | 0 | — |
case-07 | pass→pass | 26,630 | 25,803 | -3% | 1 | 1 | 0% | 4,282 | 4,677 | +9% | 0 | 0 | — |
case-08 | pass→pass | 20,705 | 21,800 | +5% | 1 | 1 | 0% | 3,150 | 3,699 | +17% | 0 | 0 | — |
case-09 | pass→pass | 19,902 | 21,278 | +7% | 1 | 1 | 0% | 2,932 | 3,777 | +29% | 0 | 0 | — |
case-10 | pass→pass | 20,534 | 24,021 | +17% | 1 | 1 | 0% | 3,116 | 3,893 | +25% | 0 | 0 | — |
case-11 | pass→pass | 21,009 | 18,165 | -14% | 1 | 1 | 0% | 3,081 | 3,138 | +2% | 0 | 0 | — |
case-12 | pass→fail | 19,167 | 19,325 | +1% | 1 | 1 | 0% | 2,853 | 3,636 | +27% | 0 | 0 | — |
case-13 | pass→pass | 19,798 | 22,081 | +12% | 1 | 1 | 0% | 3,076 | 3,991 | +30% | 0 | 0 | — |
case-14 | fail→pass | 23,912 | 23,575 | -1% | 1 | 1 | 0% | 3,418 | 3,986 | +17% | 0 | 0 | — |
case-15 | fail→pass | 20,635 | 20,135 | -2% | 1 | 1 | 0% | 3,058 | 3,582 | +17% | 0 | 0 | — |
case-16 | pass→pass | 14,662 | 14,154 | -3% | 1 | 1 | 0% | 2,252 | 2,713 | +20% | 0 | 0 | — |
case-17 | pass→pass | 16,976 | 20,914 | +23% | 1 | 1 | 0% | 2,687 | 3,777 | +41% | 0 | 0 | — |
case-18 | fail→pass | 21,254 | 22,050 | +4% | 1 | 1 | 0% | 3,187 | 3,840 | +20% | 0 | 0 | — |
case-19 | pass→pass | 21,673 | 23,233 | +7% | 1 | 1 | 0% | 3,534 | 4,217 | +19% | 0 | 0 | — |
case-20 | pass→pass | 19,139 | 18,753 | -2% | 1 | 1 | 0% | 3,342 | 3,860 | +15% | 0 | 0 | — |
case-21 | pass→pass | 15,672 | 16,440 | +5% | 1 | 1 | 0% | 2,597 | 3,147 | +21% | 0 | 0 | — |
case-22 | pass→pass | 22,527 | 23,919 | +6% | 1 | 1 | 0% | 3,364 | 4,217 | +25% | 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.