Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →银行经营分析会汇报摘要生成。当用户要求生成经营分析会材料、行长办公会简报、管理层汇报摘要、经营情况综合报告时使用。结论先行、数据支撑、风险提示的标准化输出。
.claude/skills/aifinlab-business-analysis-summary/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 33% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 9% | 0% |
银行各条线、各维度的经营分析结果分散在不同报表和分析输出中，管理层需要一份结构清晰、结论先行的综合摘要。传统做法依赖人工汇总编写，耗时且容易遗漏关键信息或重点不突出。本技能将上游各分析技能的输出整合为标准化的经营分析摘要，确保关键结论前置、数据支撑充分、风险提示明确。
| 输入项 | 说明 | 是否必须 | |--------|------|----------| | 分析结果集 | 上游各技能输出的分析结论和数据（归因、对标、结构等） | 必须 | | 摘要类型 | 综合摘要/专项摘要/简报 | 可选，默认综合摘要 | | 报告期间 | 摘要覆盖的时间范围 | 必须 | | 受众层级 | 高管层/中层/条线负责人 | 可选，默认高管层 | | 重点关注项 | 管理层特别关注的指标或领域 | 可选 |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 22,750 | 55,528 | +144% | 1 | 1 | 0% | 3,361 | 4,769 | +42% | 0 | 0 | — |
case-02 | fail→pass | 22,754 | 24,145 | +6% | 1 | 1 | 0% | 3,415 | 4,674 | +37% | 0 | 0 | — |
case-03 | fail→pass | 17,339 | 46,694 | +169% | 1 | 1 | 0% | 2,630 | 3,508 | +33% | 0 | 0 | — |
case-04 | fail→pass | 21,769 | 16,039 | -26% | 1 | 1 | 0% | 3,570 | 3,312 | -7% | 0 | 0 | — |
case-05 | fail→pass | 20,500 | 16,229 | -21% | 1 | 1 | 0% | 2,969 | 3,135 | +6% | 0 | 0 | — |
case-06 | fail→pass | 24,388 | 23,830 | -2% | 1 | 1 | 0% | 4,348 | 4,736 | +9% | 0 | 0 | — |
case-07 | pass→pass | 16,361 | 15,713 | -4% | 1 | 1 | 0% | 2,634 | 3,867 | +47% | 0 | 0 | — |
case-08 | fail→pass | 15,332 | 14,305 | -7% | 1 | 1 | 0% | 2,329 | 3,034 | +30% | 0 | 0 | — |
case-09 | pass→pass | 17,404 | 13,738 | -21% | 1 | 1 | 0% | 2,682 | 3,298 | +23% | 0 | 0 | — |
case-10 | pass→pass | 15,925 | 4,548 | -71% | 1 | 1 | 0% | 2,237 | 1,665 | -26% | 0 | 0 | — |
case-11 | fail→pass | 12,125 | 4,022 | -67% | 1 | 1 | 0% | 1,763 | 1,668 | -5% | 0 | 0 | — |
case-12 | pass→pass | 8,994 | 17,879 | +99% | 1 | 1 | 0% | 1,492 | 3,890 | +161% | 0 | 0 | — |
case-13 | pass→pass | 24,694 | 19,745 | -20% | 1 | 1 | 0% | 3,520 | 4,016 | +14% | 0 | 0 | — |
case-14 | pass→pass | 12,955 | 14,848 | +15% | 1 | 1 | 0% | 1,954 | 3,316 | +70% | 0 | 0 | — |
case-15 | pass→pass | 16,810 | 12,160 | -28% | 1 | 1 | 0% | 2,326 | 2,809 | +21% | 0 | 0 | — |
case-16 | pass→pass | 18,432 | 3,949 | -79% | 1 | 1 | 0% | 2,731 | 1,724 | -37% | 0 | 0 | — |
case-17 | fail→pass | 21,005 | 7,319 | -65% | 1 | 1 | 0% | 3,136 | 2,103 | -33% | 0 | 0 | — |
case-18 | pass→pass | 19,100 | 17,362 | -9% | 1 | 1 | 0% | 2,855 | 3,526 | +24% | 0 | 0 | — |
case-19 | pass→pass | 17,981 | 17,791 | -1% | 1 | 1 | 0% | 2,406 | 3,820 | +59% | 0 | 0 | — |
case-20 | pass→pass | 21,029 | 20,135 | -4% | 1 | 1 | 0% | 3,013 | 4,152 | +38% | 0 | 0 | — |
case-21 | fail→pass | 17,978 | 9,388 | -48% | 1 | 1 | 0% | 2,701 | 2,247 | -17% | 0 | 0 | — |
case-22 | pass→pass | 12,544 | 7,318 | -42% | 1 | 1 | 0% | 1,743 | 2,153 | +24% | 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 +41 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.