Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when wealth advisors need to craft or compare systematic investment plan (SIP) advice in a bank wealth-management context, including target setting, product selection, allocation logic, and compliant communication notes.
.claude/skills/aifinlab-bank-t188-wealth-management-sip-advice-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-21 | ✗→✓ | ▲ Improved | 46% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 78% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 60% | 0% |
本技能服务于银行财富管理中的“定投建议”场景，目标是把客户目标、风险承受能力、期限与流动性约束、产品池要素和市场环境，组织成可解释、可沟通、可合规落地的定投方案。适用于理财经理、财富顾问、私行团队在方案设计、主备方案比较、风险提示与后续跟进中使用。
scripts/run_skill.py：运行定投建议生成器，输出 Markdown/JSONscripts/sip_plan_engine.py：定投方案评分、选品与内容生成逻辑示例命令：
bashpython scripts/run_skill.py --input assets/example-input.json --format markdown
references/input-schema.mdreferences/output-schema.mdreferences/sip-plan-checklist.md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | fail→pass | 17,703 | 24,213 | +37% | 1 | 1 | 0% | 2,846 | 4,149 | +46% | 0 | 0 | — |
case-01 | fail→fail | 29,882 | 40,042 | +34% | 1 | 1 | 0% | 4,047 | 5,035 | +24% | 0 | 0 | — |
case-02 | fail→fail | 32,133 | 34,259 | +7% | 1 | 1 | 0% | 4,369 | 5,414 | +24% | 0 | 0 | — |
case-03 | pass→pass | 11,586 | 13,424 | +16% | 1 | 1 | 0% | 1,805 | 2,626 | +45% | 0 | 0 | — |
case-04 | pass→pass | 16,887 | 19,091 | +13% | 1 | 1 | 0% | 2,330 | 3,977 | +71% | 0 | 0 | — |
case-05 | fail→pass | 24,620 | 21,231 | -14% | 1 | 1 | 0% | 4,487 | 3,593 | -20% | 0 | 0 | — |
case-06 | pass→pass | 17,846 | 21,513 | +21% | 1 | 1 | 0% | 2,667 | 4,338 | +63% | 0 | 0 | — |
case-07 | pass→pass | 14,601 | 18,596 | +27% | 1 | 1 | 0% | 2,106 | 3,773 | +79% | 0 | 0 | — |
case-08 | pass→pass | 18,364 | 18,092 | -1% | 1 | 1 | 0% | 2,421 | 3,576 | +48% | 0 | 0 | — |
case-09 | pass→pass | 17,592 | 22,830 | +30% | 1 | 1 | 0% | 3,053 | 3,996 | +31% | 0 | 0 | — |
case-10 | pass→pass | 18,563 | 20,750 | +12% | 1 | 1 | 0% | 2,682 | 4,354 | +62% | 0 | 0 | — |
case-11 | fail→fail | 19,537 | 15,836 | -19% | 1 | 1 | 0% | 2,398 | 3,298 | +38% | 0 | 0 | — |
case-12 | fail→pass | 26,868 | 25,342 | -6% | 1 | 1 | 0% | 3,382 | 4,835 | +43% | 0 | 0 | — |
case-13 | fail→pass | 20,198 | 64,732 | +220% | 1 | 1 | 0% | 2,781 | 4,945 | +78% | 0 | 0 | — |
case-14 | pass→pass | 23,525 | 21,808 | -7% | 1 | 1 | 0% | 2,806 | 3,731 | +33% | 0 | 0 | — |
case-15 | fail→pass | 15,864 | 22,839 | +44% | 1 | 1 | 0% | 2,574 | 4,127 | +60% | 0 | 0 | — |
case-16 | pass→pass | 21,250 | 14,916 | -30% | 1 | 1 | 0% | 2,661 | 3,148 | +18% | 0 | 0 | — |
case-17 | pass→pass | 14,865 | 22,732 | +53% | 1 | 1 | 0% | 2,612 | 4,891 | +87% | 0 | 0 | — |
case-18 | fail→pass | 21,239 | 27,117 | +28% | 1 | 1 | 0% | 2,956 | 4,459 | +51% | 0 | 0 | — |
case-19 | pass→pass | 23,527 | 24,336 | +3% | 1 | 1 | 0% | 2,999 | 4,200 | +40% | 0 | 0 | — |
case-20 | pass→pass | 18,415 | 21,142 | +15% | 1 | 1 | 0% | 2,527 | 3,874 | +53% | 0 | 0 | — |
case-22 | pass→pass | 19,025 | 24,427 | +28% | 1 | 1 | 0% | 2,637 | 4,160 | +58% | 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 +27 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.