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，围绕贷后监测进行持续监测、风险扫描、异常识别或预警提示时使用本技能。适合输出风险信号摘要、优先级判断、处置建议和升级路径。
.claude/skills/aifinlab-bank-t198-risk-management-post-loan-monitor-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 77% | 0% |
本技能面向银行风险管理场景，目标是把预警信号、规则命中、趋势变化和处置建议串成可跟进的风险视图，而不是只给一串指标或告警日志。 当前这支 skill 更偏向服务 风控与预警团队，输出时要特别注意 把规则命中和监测结果转成业务可执行动作。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-20 | fail→fail | 25,257 | 47,673 | +89% | 1 | 1 | 0% | 2,995 | 3,690 | +23% | 0 | 0 | — |
case-01 | fail→fail | 40,568 | 31,245 | -23% | 1 | 1 | 0% | 3,783 | 4,636 | +23% | 0 | 0 | — |
case-02 | pass→pass | 24,200 | 24,395 | +1% | 1 | 1 | 0% | 3,033 | 4,405 | +45% | 0 | 0 | — |
case-03 | fail→pass | 25,986 | 19,141 | -26% | 1 | 1 | 0% | 3,542 | 3,607 | +2% | 0 | 0 | — |
case-04 | pass→pass | 21,962 | 30,212 | +38% | 1 | 1 | 0% | 3,042 | 4,558 | +50% | 0 | 0 | — |
case-05 | fail→fail | 27,326 | 22,859 | -16% | 1 | 1 | 0% | 3,490 | 4,011 | +15% | 0 | 0 | — |
case-06 | fail→pass | 23,560 | 24,934 | +6% | 1 | 1 | 0% | 2,797 | 4,021 | +44% | 0 | 0 | — |
case-07 | pass→pass | 23,578 | 24,158 | +2% | 1 | 1 | 0% | 3,026 | 4,216 | +39% | 0 | 0 | — |
case-08 | fail→pass | 29,918 | 25,134 | -16% | 1 | 1 | 0% | 2,797 | 4,122 | +47% | 0 | 0 | — |
case-09 | pass→fail | 27,182 | 25,747 | -5% | 1 | 1 | 0% | 3,115 | 4,382 | +41% | 0 | 0 | — |
case-10 | fail→fail | 37,256 | 24,650 | -34% | 1 | 1 | 0% | 3,112 | 4,210 | +35% | 0 | 0 | — |
case-11 | fail→pass | 23,341 | 28,498 | +22% | 1 | 1 | 0% | 2,896 | 4,198 | +45% | 0 | 0 | — |
case-12 | fail→fail | 23,211 | 28,079 | +21% | 1 | 1 | 0% | 3,105 | 4,222 | +36% | 0 | 0 | — |
case-13 | fail→pass | 18,986 | 23,846 | +26% | 1 | 1 | 0% | 2,404 | 4,260 | +77% | 0 | 0 | — |
case-14 | fail→pass | 26,966 | 23,098 | -14% | 1 | 1 | 0% | 3,194 | 3,684 | +15% | 0 | 0 | — |
case-15 | fail→fail | 19,065 | 21,913 | +15% | 1 | 1 | 0% | 2,718 | 3,472 | +28% | 0 | 0 | — |
case-16 | fail→pass | 19,383 | 23,417 | +21% | 1 | 1 | 0% | 2,784 | 4,061 | +46% | 0 | 0 | — |
case-17 | fail→pass | 23,211 | 30,487 | +31% | 1 | 1 | 0% | 2,995 | 5,195 | +73% | 0 | 0 | — |
case-18 | fail→pass | 17,859 | 25,200 | +41% | 1 | 1 | 0% | 2,313 | 3,966 | +71% | 0 | 0 | — |
case-19 | fail→fail | 19,215 | 23,516 | +22% | 1 | 1 | 0% | 2,618 | 4,025 | +54% | 0 | 0 | — |
case-21 | pass→fail | 21,681 | 30,075 | +39% | 1 | 1 | 0% | 2,934 | 4,379 | +49% | 0 | 0 | — |
case-22 | fail→fail | 21,538 | 19,631 | -9% | 1 | 1 | 0% | 3,062 | 3,605 | +18% | 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. 2 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.