Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当用户需要在银行零售金融场景下识别交叉销售机会并输出推荐清单时使用本技能。适合输出推荐产品、匹配理由、触达建议与合规提示。
.claude/skills/aifinlab-bank-t175-retail-finance-cross-sell-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 40% | 0% |
在零售金融场景中识别客户产品缺口与需求标签，输出可触达的交叉销售机会清单与推荐理由，支持客户经理与运营执行。
batch_id、time_windowcustomers[]（建议字段 customer_id、name、segment、products_held、need_tags、consent_marketing、churn_risk、salary_inflow）脚本入口：scripts/run_skill.py，调用 shared/retail_service_skill_engine.py 的 t175 场景。
bashpython scripts/run_skill.py --input input.json --format markdown python scripts/run_skill.py --input input.json --format json
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 27,641 | 50,738 | +84% | 1 | 1 | 0% | 3,926 | 3,921 | -0% | 0 | 0 | — |
case-02 | fail→fail | 22,199 | 27,028 | +22% | 1 | 1 | 0% | 3,061 | 3,845 | +26% | 0 | 0 | — |
case-03 | fail→pass | 19,626 | 20,064 | +2% | 1 | 1 | 0% | 2,936 | 3,288 | +12% | 0 | 0 | — |
case-04 | pass→pass | 16,021 | 17,657 | +10% | 1 | 1 | 0% | 2,351 | 2,956 | +26% | 0 | 0 | — |
case-05 | pass→pass | 16,854 | 21,949 | +30% | 1 | 1 | 0% | 2,656 | 3,412 | +28% | 0 | 0 | — |
case-06 | fail→pass | 15,091 | 14,210 | -6% | 1 | 1 | 0% | 1,833 | 2,621 | +43% | 0 | 0 | — |
case-07 | fail→pass | 27,680 | 23,657 | -15% | 1 | 1 | 0% | 3,720 | 3,718 | -0% | 0 | 0 | — |
case-08 | pass→pass | 19,317 | 17,085 | -12% | 1 | 1 | 0% | 2,584 | 2,536 | -2% | 0 | 0 | — |
case-09 | fail→fail | 34,898 | 49,459 | +42% | 1 | 1 | 0% | 4,993 | 6,374 | +28% | 0 | 0 | — |
case-10 | fail→fail | 16,313 | 16,933 | +4% | 1 | 1 | 0% | 2,390 | 2,792 | +17% | 0 | 0 | — |
case-11 | pass→pass | 20,504 | 21,421 | +4% | 1 | 1 | 0% | 2,918 | 3,373 | +16% | 0 | 0 | — |
case-12 | fail→pass | 15,785 | 16,913 | +7% | 1 | 1 | 0% | 2,469 | 3,130 | +27% | 0 | 0 | — |
case-13 | pass→pass | 16,488 | 19,076 | +16% | 1 | 1 | 0% | 2,126 | 2,843 | +34% | 0 | 0 | — |
case-14 | pass→pass | 11,297 | 16,803 | +49% | 1 | 1 | 0% | 1,660 | 3,279 | +98% | 0 | 0 | — |
case-15 | pass→pass | 20,402 | 16,656 | -18% | 1 | 1 | 0% | 2,585 | 2,890 | +12% | 0 | 0 | — |
case-16 | pass→pass | 19,972 | 16,484 | -17% | 1 | 1 | 0% | 2,687 | 2,744 | +2% | 0 | 0 | — |
case-17 | fail→pass | 15,246 | 17,551 | +15% | 1 | 1 | 0% | 2,387 | 3,338 | +40% | 0 | 0 | — |
case-18 | fail→pass | 8,932 | 5,268 | -41% | 1 | 1 | 0% | 1,589 | 1,229 | -23% | 0 | 0 | — |
case-19 | pass→pass | 21,141 | 21,650 | +2% | 1 | 1 | 0% | 2,773 | 3,245 | +17% | 0 | 0 | — |
case-20 | pass→pass | 20,252 | 18,136 | -10% | 1 | 1 | 0% | 2,854 | 3,282 | +15% | 0 | 0 | — |
case-21 | fail→pass | 22,702 | 17,947 | -21% | 1 | 1 | 0% | 2,572 | 3,206 | +25% | 0 | 0 | — |
case-22 | pass→pass | 23,352 | 13,701 | -41% | 1 | 1 | 0% | 2,393 | 2,740 | +15% | 0 | 0 | — |
case-23 | pass→pass | 16,134 | 20,979 | +30% | 1 | 1 | 0% | 2,256 | 3,653 | +62% | 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.