Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →分析各渠道交易量、成功率、交易结构与趋势变化，辅助渠道经营与资源配置决策
.claude/skills/aifinlab-channel-transaction-performance/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-16 | ✗→✓ | ▲ Improved | 76% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 57% | 0% |
银行拥有多条交易渠道，各渠道在交易量、成功率、客群结构、交易时段分布等方面表现各异。业务部门需要及时掌握各渠道交易表现，识别渠道经营中的薄弱环节，为渠道优化、资源配置和客户迁移提供数据支撑。
核心问题包括：
| 字段 | 说明 | 是否必需 | |------|------|----------| | 分析时段 | 日/周/月/季/年 | 是 | | 渠道范围 | 手机银行、网上银行、微信银行、快捷支付、开放银行API、柜面、ATM、POS、IVR、短信、代理渠道等 | 是 | | 交易类型 | 转账、支付、缴费、理财等 | 否 | | 客群维度 | 零售/对公/特定客群 | 否 | | 对比基准 | 同比/环比/指定基准期 | 否 |
bank-calc-utils：提供交易量汇总、成功率计算、同比环比等基础计算能力business-analysis-summary：将渠道交易分析结果整合至业务分析报告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 | 24,646 | 28,379 | +15% | 1 | 1 | 0% | 3,527 | 5,085 | +44% | 0 | 0 | — |
case-02 | fail→fail | 27,234 | 32,742 | +20% | 1 | 1 | 0% | 4,314 | 6,254 | +45% | 0 | 0 | — |
case-16 | fail→pass | 19,185 | 25,925 | +35% | 1 | 1 | 0% | 2,729 | 4,804 | +76% | 0 | 0 | — |
case-17 | fail→pass | 10,654 | 11,147 | +5% | 1 | 1 | 0% | 2,154 | 3,000 | +39% | 0 | 0 | — |
case-03 | fail→pass | 31,510 | 20,476 | -35% | 1 | 1 | 0% | 4,496 | 3,966 | -12% | 0 | 0 | — |
case-04 | pass→pass | 19,041 | 13,940 | -27% | 1 | 1 | 0% | 2,641 | 3,062 | +16% | 0 | 0 | — |
case-05 | pass→pass | 13,998 | 13,944 | -0% | 1 | 1 | 0% | 1,885 | 3,054 | +62% | 0 | 0 | — |
case-06 | fail→pass | 17,336 | 17,056 | -2% | 1 | 1 | 0% | 3,470 | 4,418 | +27% | 0 | 0 | — |
case-07 | fail→pass | 16,057 | 21,136 | +32% | 1 | 1 | 0% | 2,747 | 4,310 | +57% | 0 | 0 | — |
case-08 | fail→pass | 15,674 | 17,330 | +11% | 1 | 1 | 0% | 2,384 | 3,658 | +53% | 0 | 0 | — |
case-09 | fail→pass | 18,237 | 18,586 | +2% | 1 | 1 | 0% | 2,570 | 3,716 | +45% | 0 | 0 | — |
case-10 | fail→fail | 19,315 | 18,599 | -4% | 1 | 1 | 0% | 3,086 | 3,941 | +28% | 0 | 0 | — |
case-11 | pass→pass | 21,794 | 24,565 | +13% | 1 | 1 | 0% | 3,340 | 4,610 | +38% | 0 | 0 | — |
case-12 | fail→pass | 17,831 | 19,430 | +9% | 1 | 1 | 0% | 2,864 | 4,013 | +40% | 0 | 0 | — |
case-13 | fail→fail | 16,258 | 20,317 | +25% | 1 | 1 | 0% | 2,449 | 4,121 | +68% | 0 | 0 | — |
case-14 | fail→fail | 19,696 | 23,415 | +19% | 1 | 1 | 0% | 2,970 | 4,421 | +49% | 0 | 0 | — |
case-15 | pass→pass | 19,766 | 16,699 | -16% | 1 | 1 | 0% | 3,369 | 3,725 | +11% | 0 | 0 | — |
case-18 | fail→pass | 16,323 | 19,793 | +21% | 1 | 1 | 0% | 2,483 | 3,992 | +61% | 0 | 0 | — |
case-19 | fail→pass | 15,034 | 17,657 | +17% | 1 | 1 | 0% | 2,595 | 4,120 | +59% | 0 | 0 | — |
case-20 | fail→pass | 9,976 | 13,643 | +37% | 1 | 1 | 0% | 1,595 | 3,159 | +98% | 0 | 0 | — |
case-21 | pass→pass | 19,678 | 21,460 | +9% | 1 | 1 | 0% | 2,847 | 4,217 | +48% | 0 | 0 | — |
case-22 | fail→pass | 19,186 | 16,975 | -12% | 1 | 1 | 0% | 2,743 | 3,831 | +40% | 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 +55 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.