Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →用于银行零售金融场景下的信用卡营销机会识别任务。适用于基于客户画像、资产负债、 交易行为、渠道互动、生命周期阶段、权益偏好以及外部行业数据，识别潜在办卡、 升额、分期、附属卡、权益升级、激活促活与交叉营销机会。优先级为 P1。 当用户需要对零售客户进行营销推荐、名单筛选、机会评分、活动匹配、客群策略制定、 信用卡经营分析或机会归因时，应使用本技能。 本技能对行业数据依赖较强：需要参考市场渗透率、同业产品供给、行业活跃水平、 客群偏好基准、渠道转化基准、地区与客群差异基准等外部或行业数据。
.claude/skills/aifinlab-credit-card-marketing-opportunity-identification-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 48% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 76% | 0% |
| case-26 | ✗→✓ | ▲ Improved | 56% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 99% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 139% | 0% |
本技能用于识别零售客户在信用卡业务中的营销机会，包括但不限于：
本技能输出的核心不是“直接给客户推什么”，而是：
在以下场景下优先使用本技能：
遇到以下情况时，应降低结论强度，并明确标注“待补充行业数据”或“仅为内部信号提示”：
优先输入如下信息：
输出时应至少包含以下部分：
重点识别以下信号：
重点识别以下信号：
可识别的场景包括：
建议根据客户行为偏好选择渠道：
本技能需要较强行业数据支持。应优先使用以下外部或行业数据：
若缺少上述数据，则：
references/marketing_opportunity_methodology.mdreferences/industry_data_requirements.mdreferences/customer_product_mapping.mdreferences/compliance_notes.mdreferences/output_schema.mdscripts/card_opportunity_signal_engine.py：信用卡营销机会信号识别scripts/industry_benchmark_calibrator.py：行业基准校准scripts/render_marketing_report.py：渲染营销机会识别报告| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 23,222 | 51,013 | +120% | 1 | 1 | 0% | 3,223 | 4,774 | +48% | 0 | 0 | — |
case-02 | pass→pass | 15,177 | 18,486 | +22% | 1 | 1 | 0% | 2,264 | 4,510 | +99% | 0 | 0 | — |
case-03 | pass→pass | 17,656 | 25,342 | +44% | 1 | 1 | 0% | 2,384 | 5,706 | +139% | 0 | 0 | — |
case-04 | pass→pass | 18,213 | 23,771 | +31% | 1 | 1 | 0% | 2,527 | 5,108 | +102% | 0 | 0 | — |
case-05 | pass→pass | 18,580 | 23,699 | +28% | 1 | 1 | 0% | 2,576 | 5,180 | +101% | 0 | 0 | — |
case-10 | pass→pass | 17,161 | 22,568 | +32% | 1 | 1 | 0% | 2,495 | 5,132 | +106% | 0 | 0 | — |
case-17 | pass→pass | 18,724 | 17,313 | -8% | 1 | 1 | 0% | 2,648 | 4,183 | +58% | 0 | 0 | — |
case-18 | pass→pass | 18,206 | 18,291 | +0% | 1 | 1 | 0% | 2,696 | 4,559 | +69% | 0 | 0 | — |
case-27 | fail→fail | 22,450 | 36,044 | +61% | 1 | 1 | 0% | 3,873 | 7,903 | +104% | 0 | 0 | — |
case-06 | pass→pass | 12,837 | 20,842 | +62% | 1 | 1 | 0% | 1,990 | 4,861 | +144% | 0 | 0 | — |
case-07 | pass→pass | 14,028 | 27,785 | +98% | 1 | 1 | 0% | 2,120 | 5,825 | +175% | 0 | 0 | — |
case-08 | pass→pass | 18,219 | 24,833 | +36% | 1 | 1 | 0% | 2,655 | 5,345 | +101% | 0 | 0 | — |
case-09 | pass→pass | 32,214 | 21,977 | -32% | 1 | 1 | 0% | 2,844 | 5,117 | +80% | 0 | 0 | — |
case-11 | pass→pass | 21,172 | 35,005 | +65% | 1 | 1 | 0% | 2,910 | 6,882 | +136% | 0 | 0 | — |
case-12 | pass→pass | 17,198 | 17,913 | +4% | 1 | 1 | 0% | 2,294 | 4,330 | +89% | 0 | 0 | — |
case-13 | fail→pass | 17,578 | 20,558 | +17% | 1 | 1 | 0% | 2,738 | 4,811 | +76% | 0 | 0 | — |
case-14 | pass→pass | 13,532 | 19,393 | +43% | 1 | 1 | 0% | 2,017 | 4,645 | +130% | 0 | 0 | — |
case-15 | pass→pass | 14,030 | 21,029 | +50% | 1 | 1 | 0% | 1,999 | 4,847 | +142% | 0 | 0 | — |
case-16 | pass→pass | 19,393 | 20,433 | +5% | 1 | 1 | 0% | 2,676 | 4,643 | +74% | 0 | 0 | — |
case-19 | pass→pass | 16,915 | 26,015 | +54% | 1 | 1 | 0% | 2,479 | 5,645 | +128% | 0 | 0 | — |
case-20 | pass→pass | 17,245 | 19,264 | +12% | 1 | 1 | 0% | 2,393 | 4,496 | +88% | 0 | 0 | — |
case-21 | pass→pass | 15,915 | 24,601 | +55% | 1 | 1 | 0% | 2,152 | 5,131 | +138% | 0 | 0 | — |
case-22 | pass→pass | 19,947 | 23,015 | +15% | 1 | 1 | 0% | 2,727 | 4,867 | +78% | 0 | 0 | — |
case-23 | pass→pass | 19,405 | 30,631 | +58% | 1 | 1 | 0% | 2,757 | 6,077 | +120% | 0 | 0 | — |
case-24 | pass→pass | 19,335 | 25,771 | +33% | 1 | 1 | 0% | 2,790 | 5,588 | +100% | 0 | 0 | — |
case-25 | pass→pass | 12,545 | 16,289 | +30% | 1 | 1 | 0% | 1,738 | 4,056 | +133% | 0 | 0 | — |
case-26 | fail→pass | 25,808 | 26,468 | +3% | 1 | 1 | 0% | 3,536 | 5,511 | +56% | 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. 27 cases were attempted. The headline lift of +11 percentage points is the difference between those two pass rates over the 27 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.