Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当需要生成风险管理周报并回顾一周风险趋势、重点事件和处置进展时使用本技能。
.claude/skills/aifinlab-bank-t214-risk-management-weekly-report-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -64% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -54% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -45% | 0% |
用于每周汇总风险信号趋势、重点事件与处置进展，输出结构化周报内容，帮助风控团队复盘与计划下一周期。
scripts/weekly_risk_report_builder.py：生成周度风险报告草稿。| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,965 | 12,008 | -43% | 1 | 1 | 0% | 2,625 | 1,870 | -29% | 0 | 0 | — |
case-02 | pass→pass | 21,805 | 11,786 | -46% | 1 | 1 | 0% | 2,685 | 1,758 | -35% | 0 | 0 | — |
case-03 | fail→pass | 22,653 | 13,171 | -42% | 1 | 1 | 0% | 3,119 | 2,011 | -36% | 0 | 0 | — |
case-04 | fail→pass | 11,219 | 1,848 | -84% | 1 | 1 | 0% | 1,471 | 533 | -64% | 0 | 0 | — |
case-05 | pass→pass | 15,006 | 15,406 | +3% | 1 | 1 | 0% | 2,279 | 2,252 | -1% | 0 | 0 | — |
case-06 | pass→pass | 17,517 | 15,081 | -14% | 1 | 1 | 0% | 2,090 | 2,235 | +7% | 0 | 0 | — |
case-07 | pass→pass | 16,818 | 16,417 | -2% | 1 | 1 | 0% | 2,473 | 2,411 | -3% | 0 | 0 | — |
case-08 | fail→pass | 10,496 | 4,411 | -58% | 1 | 1 | 0% | 1,707 | 787 | -54% | 0 | 0 | — |
case-09 | pass→pass | 20,982 | 11,893 | -43% | 1 | 1 | 0% | 2,472 | 2,078 | -16% | 0 | 0 | — |
case-10 | pass→pass | 15,845 | 14,519 | -8% | 1 | 1 | 0% | 2,259 | 2,463 | +9% | 0 | 0 | — |
case-11 | pass→pass | 20,202 | 15,175 | -25% | 1 | 1 | 0% | 2,684 | 2,263 | -16% | 0 | 0 | — |
case-12 | pass→pass | 20,610 | 20,681 | +0% | 1 | 1 | 0% | 2,646 | 2,889 | +9% | 0 | 0 | — |
case-13 | pass→pass | 18,533 | 19,358 | +4% | 1 | 1 | 0% | 2,575 | 2,771 | +8% | 0 | 0 | — |
case-14 | pass→pass | 16,842 | 14,231 | -16% | 1 | 1 | 0% | 2,139 | 2,230 | +4% | 0 | 0 | — |
case-15 | pass→pass | 20,537 | 17,563 | -14% | 1 | 1 | 0% | 2,636 | 2,604 | -1% | 0 | 0 | — |
case-16 | pass→pass | 17,456 | 16,588 | -5% | 1 | 1 | 0% | 2,135 | 2,437 | +14% | 0 | 0 | — |
case-17 | pass→pass | 12,609 | 11,716 | -7% | 1 | 1 | 0% | 1,722 | 1,789 | +4% | 0 | 0 | — |
case-18 | pass→pass | 15,491 | 12,169 | -21% | 1 | 1 | 0% | 1,864 | 1,817 | -3% | 0 | 0 | — |
case-19 | fail→pass | 6,123 | 1,410 | -77% | 1 | 1 | 0% | 972 | 538 | -45% | 0 | 0 | — |
case-20 | pass→pass | 18,211 | 15,325 | -16% | 1 | 1 | 0% | 2,251 | 2,197 | -2% | 0 | 0 | — |
case-21 | pass→pass | 16,769 | 11,201 | -33% | 1 | 1 | 0% | 1,985 | 2,005 | +1% | 0 | 0 | — |
case-22 | pass→pass | 14,125 | 28,057 | +99% | 1 | 1 | 0% | 1,832 | 946 | -48% | 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 +23 percentage points is the difference between those two pass rates over the 22 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.