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，围绕家庭配置进行适配、配置、推荐、方案设计或备选方案比较时使用本技能。适合输出主方案、备选方案、匹配逻辑、风险提示和沟通要点。
.claude/skills/aifinlab-bank-t183-wealth-management-family-allocation-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 97% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 77% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 79% | 0% |
本技能面向银行财富管理场景，核心是把客户目标、风险承受能力、期限约束、家庭资产配置和市场环境组织成可解释、可沟通、可合规落地的建议。当前这支 skill 更偏向服务理财经理、财富顾问和私行团队，输出时要特别注意把客户目标、风险偏好和市场环境组织成可解释建议，同时明确假设和边界。
最低必需字段
增强字段（有助于提高可执行性）
结构化输入建议（JSON）
json{ "name": "张三家庭", "risk_level": "稳健", "horizon_years": 5, "liquidity_months": 6, "total_assets": 5000000, "monthly_expense": 30000, "goals": [ {"name": "子女教育", "amount": 800000, "horizon_years": 4}, {"name": "养老准备", "amount": 2000000, "horizon_years": 10} ], "constraints": {"min_liquidity_ratio": 0.15, "max_growth_ratio": 0.45}, "holdings": [ {"asset": "货币类", "amount": 800000}, {"asset": "固收+产品", "amount": 1200000} ], "product_pool": [ {"name": "稳健固收+组合", "type": "income", "risk": "中低"}, {"name": "权益增强组合", "type": "growth", "risk": "中高"} ] }
必须包含
推荐输出模板
本技能包含可复用脚本，用于生成家庭资产配置草案（比例与沟通要点），便于快速形成标准化输出。
脚本位置
scripts/build_family_allocation.py适用场景
使用方式
bashpython scripts/build_family_allocation.py --input input.json --output plan.md
输出说明
--format json 输出结构化结果| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 30,828 | 36,914 | +20% | 1 | 1 | 0% | 4,621 | 5,276 | +14% | 0 | 0 | — |
case-02 | fail→fail | 30,074 | 37,764 | +26% | 1 | 1 | 0% | 4,474 | 6,488 | +45% | 0 | 0 | — |
case-03 | fail→pass | 17,493 | 19,226 | +10% | 1 | 1 | 0% | 2,060 | 4,058 | +97% | 0 | 0 | — |
case-04 | pass→pass | 16,080 | 23,293 | +45% | 1 | 1 | 0% | 2,361 | 4,222 | +79% | 0 | 0 | — |
case-05 | fail→pass | 8,107 | 2,244 | -72% | 1 | 1 | 0% | 1,440 | 1,774 | +23% | 0 | 0 | — |
case-06 | pass→pass | 6,480 | 2,543 | -61% | 1 | 1 | 0% | 1,026 | 1,725 | +68% | 0 | 0 | — |
case-07 | fail→fail | 27,650 | 34,104 | +23% | 1 | 1 | 0% | 4,246 | 6,062 | +43% | 0 | 0 | — |
case-08 | pass→pass | 21,008 | 34,267 | +63% | 1 | 1 | 0% | 3,536 | 6,477 | +83% | 0 | 0 | — |
case-09 | pass→pass | 11,940 | 12,230 | +2% | 1 | 1 | 0% | 1,621 | 3,047 | +88% | 0 | 0 | — |
case-10 | pass→pass | 15,300 | 19,441 | +27% | 1 | 1 | 0% | 2,216 | 4,138 | +87% | 0 | 0 | — |
case-11 | pass→pass | 12,907 | 18,973 | +47% | 1 | 1 | 0% | 1,935 | 4,072 | +110% | 0 | 0 | — |
case-12 | pass→pass | 10,430 | 4,296 | -59% | 1 | 1 | 0% | 1,513 | 1,908 | +26% | 0 | 0 | — |
case-13 | fail→pass | 12,549 | 10,283 | -18% | 1 | 1 | 0% | 1,602 | 2,832 | +77% | 0 | 0 | — |
case-14 | pass→pass | 10,861 | 6,565 | -40% | 1 | 1 | 0% | 1,338 | 2,142 | +60% | 0 | 0 | — |
case-15 | pass→pass | 11,594 | 5,018 | -57% | 1 | 1 | 0% | 1,507 | 1,992 | +32% | 0 | 0 | — |
case-16 | fail→fail | 19,278 | 25,969 | +35% | 1 | 1 | 0% | 2,918 | 4,477 | +53% | 0 | 0 | — |
case-17 | fail→fail | 25,796 | 35,133 | +36% | 1 | 1 | 0% | 4,006 | 5,411 | +35% | 0 | 0 | — |
case-18 | fail→fail | 22,893 | 24,879 | +9% | 1 | 1 | 0% | 3,019 | 4,658 | +54% | 0 | 0 | — |
case-19 | fail→fail | 25,596 | 25,123 | -2% | 1 | 1 | 0% | 3,308 | 4,939 | +49% | 0 | 0 | — |
case-20 | fail→pass | 19,650 | 19,473 | -1% | 1 | 1 | 0% | 2,779 | 4,432 | +59% | 0 | 0 | — |
case-21 | pass→pass | 18,085 | 19,583 | +8% | 1 | 1 | 0% | 2,740 | 4,206 | +54% | 0 | 0 | — |
case-22 | pass→pass | 15,240 | 21,757 | +43% | 1 | 1 | 0% | 2,303 | 4,331 | +88% | 0 | 0 | — |
case-23 | fail→fail | 26,988 | 26,814 | -1% | 1 | 1 | 0% | 3,376 | 4,896 | +45% | 0 | 0 | — |
case-24 | pass→pass | 19,344 | 25,663 | +33% | 1 | 1 | 0% | 2,848 | 4,148 | +46% | 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. 24 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 24 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.