Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当用户需要在银行风险管理场景下，对高风险账户进行持续监测、异常识别和处置建议输出时使用本技能。适合生成账户风险摘要、优先级排序、排查清单与升级路径。
.claude/skills/aifinlab-bank-t207-risk-management-account-monitor-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 16% | 0% |
本技能面向银行风险管理场景，聚焦账户层面的异常交易行为、账户活动强度变化和风险标签积累。目标是将规则命中、账户画像变化与历史处置记录转为可执行的排查与处置动作，支持风控、预警运营与合规团队。
scripts/account_monitoring_scan.py：用于账户异常规则命中、分级与摘要输出| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 27,340 | 36,514 | +34% | 1 | 1 | 0% | 3,912 | 4,480 | +15% | 0 | 0 | — |
case-02 | fail→pass | 28,426 | 25,451 | -10% | 1 | 1 | 0% | 3,656 | 4,236 | +16% | 0 | 0 | — |
case-03 | pass→pass | 21,185 | 18,859 | -11% | 1 | 1 | 0% | 2,565 | 3,024 | +18% | 0 | 0 | — |
case-04 | pass→pass | 16,443 | 16,425 | -0% | 1 | 1 | 0% | 1,973 | 2,783 | +41% | 0 | 0 | — |
case-05 | pass→pass | 17,119 | 19,888 | +16% | 1 | 1 | 0% | 2,225 | 2,946 | +32% | 0 | 0 | — |
case-06 | fail→pass | 19,788 | 25,629 | +30% | 1 | 1 | 0% | 2,853 | 4,338 | +52% | 0 | 0 | — |
case-07 | fail→pass | 22,977 | 27,153 | +18% | 1 | 1 | 0% | 3,294 | 3,909 | +19% | 0 | 0 | — |
case-08 | fail→pass | 22,404 | 28,959 | +29% | 1 | 1 | 0% | 2,824 | 4,301 | +52% | 0 | 0 | — |
case-09 | pass→pass | 17,895 | 23,493 | +31% | 1 | 1 | 0% | 2,372 | 3,469 | +46% | 0 | 0 | — |
case-10 | fail→pass | 22,464 | 20,740 | -8% | 1 | 1 | 0% | 2,862 | 3,327 | +16% | 0 | 0 | — |
case-11 | pass→pass | 64,654 | 20,304 | -69% | 1 | 1 | 0% | 2,894 | 3,696 | +28% | 0 | 0 | — |
case-12 | fail→fail | 22,262 | 19,237 | -14% | 1 | 1 | 0% | 2,691 | 3,563 | +32% | 0 | 0 | — |
case-13 | fail→pass | 22,942 | 23,359 | +2% | 1 | 1 | 0% | 2,855 | 3,724 | +30% | 0 | 0 | — |
case-14 | pass→pass | 21,310 | 23,327 | +9% | 1 | 1 | 0% | 2,771 | 3,528 | +27% | 0 | 0 | — |
case-15 | pass→pass | 24,729 | 24,906 | +1% | 1 | 1 | 0% | 3,234 | 4,099 | +27% | 0 | 0 | — |
case-16 | pass→fail | 20,458 | 19,864 | -3% | 1 | 1 | 0% | 2,812 | 3,427 | +22% | 0 | 0 | — |
case-17 | fail→pass | 21,862 | 22,164 | +1% | 1 | 1 | 0% | 2,913 | 3,888 | +33% | 0 | 0 | — |
case-18 | fail→pass | 20,059 | 21,644 | +8% | 1 | 1 | 0% | 2,683 | 3,279 | +22% | 0 | 0 | — |
case-19 | fail→pass | 54,606 | 56,022 | +3% | 1 | 1 | 0% | 4,105 | 4,508 | +10% | 0 | 0 | — |
case-20 | pass→pass | 19,919 | 20,025 | +1% | 1 | 1 | 0% | 2,230 | 3,308 | +48% | 0 | 0 | — |
case-21 | pass→pass | 25,009 | 21,816 | -13% | 1 | 1 | 0% | 3,412 | 3,566 | +5% | 0 | 0 | — |
case-22 | pass→pass | 19,626 | 22,504 | +15% | 1 | 1 | 0% | 2,691 | 3,439 | +28% | 0 | 0 | — |
case-23 | pass→pass | 21,371 | 21,724 | +2% | 1 | 1 | 0% | 3,052 | 3,745 | +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 +35 percentage points is the difference between those two pass rates over the 23 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.