Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →分析客户或客群的产品持有率、渗透结构和覆盖深度，识别交叉销售机会和单产品依赖风险，支撑产品营销策略制定。
.claude/skills/aifinlab-customer-product-penetration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 18% | 0% |
银行零售业务收入和客户粘性与产品渗透深度高度相关。多产品客户流失率低、贡献度高，但实际经营中大量客户仅持有单一产品，交叉销售潜力未被充分挖掘。需要系统分析产品渗透率和持有结构，识别交叉销售机会和单产品依赖风险。
| 输入项 | 说明 | 是否必填 | |--------|------|----------| | 客户列表 | 待分析的客户编号集合 | 必填 | | 产品持有明细 | 每位客户当前持有的产品清单 | 必填 | | 产品分类标准 | 产品归类规则（默认五大类） | 选填 | | 客户价值层级 | 分层结果（如有） | 选填 | | 产品到期信息 | 理财/存款/贷款到期日 | 选填 | | 历史持有记录 | 产品持有变化历史（如有） | 选填 | | 机构维度 | 分行/支行/网点归属 | 选填 |
python# 调用 skill result = run_skill({ "param1": "value1", "param2": "value2" })
bashpython scripts/run_skill.py --input data.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 | 28,376 | 33,964 | +20% | 1 | 1 | 0% | 4,464 | 6,507 | +46% | 0 | 0 | — |
case-02 | fail→fail | 26,151 | 27,400 | +5% | 1 | 1 | 0% | 3,937 | 5,465 | +39% | 0 | 0 | — |
case-03 | fail→pass | 28,339 | 25,632 | -10% | 1 | 1 | 0% | 4,516 | 5,347 | +18% | 0 | 0 | — |
case-04 | fail→pass | 21,066 | 8,531 | -60% | 1 | 1 | 0% | 2,983 | 2,262 | -24% | 0 | 0 | — |
case-05 | fail→pass | 20,348 | 10,227 | -50% | 1 | 1 | 0% | 3,167 | 2,544 | -20% | 0 | 0 | — |
case-06 | fail→pass | 21,406 | 9,458 | -56% | 1 | 1 | 0% | 3,053 | 2,361 | -23% | 0 | 0 | — |
case-07 | pass→pass | 18,940 | 16,601 | -12% | 1 | 1 | 0% | 3,079 | 3,630 | +18% | 0 | 0 | — |
case-08 | fail→pass | 21,315 | 18,585 | -13% | 1 | 1 | 0% | 3,164 | 3,725 | +18% | 0 | 0 | — |
case-09 | pass→pass | 12,487 | 10,321 | -17% | 1 | 1 | 0% | 1,855 | 2,687 | +45% | 0 | 0 | — |
case-10 | fail→pass | 16,938 | 15,827 | -7% | 1 | 1 | 0% | 2,517 | 3,431 | +36% | 0 | 0 | — |
case-11 | pass→pass | 20,349 | 19,252 | -5% | 1 | 1 | 0% | 2,967 | 3,929 | +32% | 0 | 0 | — |
case-12 | fail→pass | 19,433 | 16,666 | -14% | 1 | 1 | 0% | 2,878 | 3,479 | +21% | 0 | 0 | — |
case-13 | pass→pass | 14,697 | 11,380 | -23% | 1 | 1 | 0% | 2,116 | 2,662 | +26% | 0 | 0 | — |
case-14 | pass→pass | 20,267 | 16,165 | -20% | 1 | 1 | 0% | 2,999 | 3,378 | +13% | 0 | 0 | — |
case-15 | fail→pass | 12,954 | 8,338 | -36% | 1 | 1 | 0% | 1,930 | 2,309 | +20% | 0 | 0 | — |
case-16 | fail→pass | 15,635 | 4,228 | -73% | 1 | 1 | 0% | 2,281 | 1,626 | -29% | 0 | 0 | — |
case-17 | fail→pass | 16,136 | 2,079 | -87% | 1 | 1 | 0% | 2,388 | 1,308 | -45% | 0 | 0 | — |
case-18 | fail→pass | 15,657 | 4,447 | -72% | 1 | 1 | 0% | 2,518 | 1,797 | -29% | 0 | 0 | — |
case-19 | pass→pass | 15,812 | 3,584 | -77% | 1 | 1 | 0% | 2,404 | 1,604 | -33% | 0 | 0 | — |
case-20 | pass→pass | 15,984 | 10,715 | -33% | 1 | 1 | 0% | 2,393 | 2,534 | +6% | 0 | 0 | — |
case-21 | pass→pass | 17,773 | 10,805 | -39% | 1 | 1 | 0% | 2,824 | 2,689 | -5% | 0 | 0 | — |
case-22 | fail→fail | 15,198 | 9,004 | -41% | 1 | 1 | 0% | 2,291 | 2,409 | +5% | 0 | 0 | — |
case-23 | pass→pass | 16,364 | 12,541 | -23% | 1 | 1 | 0% | 2,237 | 2,903 | +30% | 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 +48 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.