Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当需要对反欺诈信号进行初筛、分层并输出处置建议与升级路径时使用本技能。
.claude/skills/aifinlab-bank-t219-risk-management-anti-fraud-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 2% | 0% |
用于汇总反欺诈信号与命中规则，生成初筛结论、优先级、排查清单与升级建议，支持反欺诈运营与风险排查。
scripts/anti_fraud_triage.py：生成反欺诈初筛报告草稿。| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 23,485 | 21,958 | -7% | 1 | 1 | 0% | 3,235 | 3,217 | -1% | 0 | 0 | — |
case-02 | fail→pass | 24,829 | 28,458 | +15% | 1 | 1 | 0% | 3,650 | 3,903 | +7% | 0 | 0 | — |
case-03 | pass→pass | 21,680 | 18,162 | -16% | 1 | 1 | 0% | 2,587 | 2,675 | +3% | 0 | 0 | — |
case-04 | pass→pass | 15,811 | 14,706 | -7% | 1 | 1 | 0% | 2,008 | 2,443 | +22% | 0 | 0 | — |
case-05 | fail→pass | 12,467 | 15,934 | +28% | 1 | 1 | 0% | 1,592 | 2,227 | +40% | 0 | 0 | — |
case-06 | pass→pass | 16,602 | 14,384 | -13% | 1 | 1 | 0% | 2,384 | 2,438 | +2% | 0 | 0 | — |
case-07 | fail→pass | 28,281 | 22,680 | -20% | 1 | 1 | 0% | 5,403 | 3,519 | -35% | 0 | 0 | — |
case-08 | pass→pass | 17,997 | 19,988 | +11% | 1 | 1 | 0% | 2,308 | 2,860 | +24% | 0 | 0 | — |
case-09 | pass→pass | 19,740 | 21,005 | +6% | 1 | 1 | 0% | 2,782 | 3,003 | +8% | 0 | 0 | — |
case-10 | pass→pass | 14,604 | 13,052 | -11% | 1 | 1 | 0% | 2,051 | 2,361 | +15% | 0 | 0 | — |
case-11 | pass→pass | 11,195 | 18,680 | +67% | 1 | 1 | 0% | 1,780 | 2,628 | +48% | 0 | 0 | — |
case-12 | pass→pass | 24,908 | 17,384 | -30% | 1 | 1 | 0% | 2,794 | 2,896 | +4% | 0 | 0 | — |
case-13 | pass→pass | 23,275 | 22,255 | -4% | 1 | 1 | 0% | 2,853 | 3,309 | +16% | 0 | 0 | — |
case-14 | pass→pass | 19,819 | 20,492 | +3% | 1 | 1 | 0% | 2,837 | 3,215 | +13% | 0 | 0 | — |
case-15 | pass→pass | 16,995 | 17,394 | +2% | 1 | 1 | 0% | 2,202 | 2,896 | +32% | 0 | 0 | — |
case-16 | pass→pass | 16,682 | 15,537 | -7% | 1 | 1 | 0% | 2,373 | 2,609 | +10% | 0 | 0 | — |
case-17 | fail→pass | 22,491 | 19,522 | -13% | 1 | 1 | 0% | 2,810 | 2,864 | +2% | 0 | 0 | — |
case-18 | pass→pass | 16,771 | 17,442 | +4% | 1 | 1 | 0% | 2,165 | 2,817 | +30% | 0 | 0 | — |
case-19 | pass→pass | 22,698 | 17,768 | -22% | 1 | 1 | 0% | 3,066 | 2,851 | -7% | 0 | 0 | — |
case-20 | fail→pass | 18,679 | 22,052 | +18% | 1 | 1 | 0% | 2,346 | 3,030 | +29% | 0 | 0 | — |
case-21 | pass→pass | 21,354 | 19,214 | -10% | 1 | 1 | 0% | 2,894 | 3,157 | +9% | 0 | 0 | — |
case-22 | fail→pass | 18,700 | 14,389 | -23% | 1 | 1 | 0% | 2,764 | 2,334 | -16% | 0 | 0 | — |
case-23 | pass→pass | 17,484 | 18,837 | +8% | 1 | 1 | 0% | 2,386 | 2,931 | +23% | 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. 23 cases were attempted. The headline lift of +30 percentage points is the difference between those two pass rates over the 23 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.