Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当用户需要在银行财富管理场景下制作客户陪伴月报、复盘摘要、跟进清单或月度沟通要点时使用本技能，帮助把多来源材料沉淀为结构化、可执行、可合规的月度陪伴内容。
.claude/skills/aifinlab-bank-t193-wealth-management-customer-companion-monthly-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-21 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-23 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-11 | ✓→✗ | ▼ Worse | 40% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 109% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 104% | 0% |
本技能面向银行财富管理场景，目标是将客户目标、风险承受能力、期限约束、家庭资产配置和市场环境沉淀为可解释、可沟通、可合规落地的月度陪伴报告。重点服务客户陪伴运营人员、理财经理与财富顾问，输出需突出月度复盘、组合回顾、风险说明与下月关注点。
scripts/monthly_companion_report_builder.py 进行字段标准化与模板生成如需批量生成或统一输出结构，可使用脚本进行输入规范化与模板化输出。
scripts/monthly_companion_report_builder.py| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 11,468 | 15,066 | +31% | 1 | 1 | 0% | 1,617 | 3,385 | +109% | 0 | 0 | — |
case-02 | pass→pass | 10,947 | 19,152 | +75% | 1 | 1 | 0% | 1,755 | 3,581 | +104% | 0 | 0 | — |
case-03 | pass→pass | 8,845 | 11,114 | +26% | 1 | 1 | 0% | 1,294 | 2,988 | +131% | 0 | 0 | — |
case-04 | pass→pass | 12,056 | 19,257 | +60% | 1 | 1 | 0% | 1,721 | 3,674 | +113% | 0 | 0 | — |
case-05 | pass→pass | 13,365 | 16,919 | +27% | 1 | 1 | 0% | 1,705 | 3,293 | +93% | 0 | 0 | — |
case-06 | pass→pass | 12,308 | 17,376 | +41% | 1 | 1 | 0% | 1,631 | 3,203 | +96% | 0 | 0 | — |
case-07 | pass→pass | 14,555 | 17,120 | +18% | 1 | 1 | 0% | 1,871 | 3,442 | +84% | 0 | 0 | — |
case-08 | pass→pass | 11,128 | 14,705 | +32% | 1 | 1 | 0% | 1,748 | 3,485 | +99% | 0 | 0 | — |
case-09 | pass→pass | 11,354 | 14,626 | +29% | 1 | 1 | 0% | 1,665 | 3,251 | +95% | 0 | 0 | — |
case-10 | pass→pass | 10,142 | 12,993 | +28% | 1 | 1 | 0% | 1,605 | 2,779 | +73% | 0 | 0 | — |
case-11 | pass→fail | 11,255 | 7,407 | -34% | 1 | 1 | 0% | 1,476 | 2,062 | +40% | 0 | 0 | — |
case-12 | pass→pass | 15,080 | 16,823 | +12% | 1 | 1 | 0% | 1,996 | 3,597 | +80% | 0 | 0 | — |
case-13 | pass→pass | 9,002 | 17,536 | +95% | 1 | 1 | 0% | 1,254 | 3,758 | +200% | 0 | 0 | — |
case-14 | pass→pass | 11,605 | 23,400 | +102% | 1 | 1 | 0% | 1,838 | 3,569 | +94% | 0 | 0 | — |
case-15 | fail→fail | 4,790 | 4,976 | +4% | 1 | 1 | 0% | 571 | 1,800 | +215% | 0 | 0 | — |
case-16 | fail→fail | 4,190 | 4,116 | -2% | 1 | 1 | 0% | 612 | 1,673 | +173% | 0 | 0 | — |
case-17 | pass→pass | 11,998 | 17,975 | +50% | 1 | 1 | 0% | 1,579 | 2,954 | +87% | 0 | 0 | — |
case-18 | pass→pass | 14,509 | 16,876 | +16% | 1 | 1 | 0% | 1,858 | 3,112 | +67% | 0 | 0 | — |
case-19 | pass→pass | 7,031 | 7,624 | +8% | 1 | 1 | 0% | 1,035 | 1,999 | +93% | 0 | 0 | — |
case-20 | pass→pass | 8,771 | 6,341 | -28% | 1 | 1 | 0% | 1,268 | 1,998 | +58% | 0 | 0 | — |
case-21 | fail→pass | 15,878 | 13,013 | -18% | 1 | 1 | 0% | 2,467 | 2,755 | +12% | 0 | 0 | — |
case-22 | pass→pass | 9,219 | 14,389 | +56% | 1 | 1 | 0% | 1,155 | 2,927 | +153% | 0 | 0 | — |
case-23 | fail→pass | 23,288 | 9,789 | -58% | 1 | 1 | 0% | 3,316 | 2,281 | -31% | 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 +4 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.