Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →生成再融资尽职调查问题清单，覆盖前次募资使用、本次募资必要性等
.claude/skills/aifinlab-9b0338/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 46% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 39% | 0% |
本技能用于生成再融资（定增、配股、可转债等）尽职调查问题清单，重点关注前次募集资金使用、本次募资必要性及可行性。
选择再融资类型（定增/配股/可转债）
bashcd "C:\Users\51994\Desktop\skills\证券 2\15. 尽调问题清单助手 - 再融资版" python scripts/generate_refinance_dd.py --type <融资类型> --output <输出路径>
问题清单将按领域分类，包含具体问题
根据问题清单开展尽职调查
markdown## 再融资尽职调查问题清单 ### 融资类型：定向增发 ### 一、公司基本情况 ... ### 二、前次募集资金使用 #### 3. 前次募资基本情况 3.1 请说明前次募集资金的时间、方式、金额 3.2 请说明前次募投项目情况 3.3 请提供前次募资验资报告 ... ### 三、本次募集资金必要性 ... ### 四、本次募投项目 ... ### 五、财务情况 ... ### 六、法律合规 ... ### 七、发行方案 ...
scripts/generate_refinance_dd.py - 问题清单生成脚本references/refinance_checklist_template.md - 再融资尽调清单模板| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 23,459 | 40,048 | +71% | 1 | 1 | 0% | 3,640 | 5,596 | +54% | 0 | 0 | — |
case-02 | fail→fail | 24,171 | 35,456 | +47% | 1 | 1 | 0% | 3,722 | 6,794 | +83% | 0 | 0 | — |
case-03 | fail→fail | 25,642 | 24,580 | -4% | 1 | 1 | 0% | 3,831 | 5,138 | +34% | 0 | 0 | — |
case-04 | pass→pass | 27,092 | 25,118 | -7% | 1 | 1 | 0% | 4,171 | 5,282 | +27% | 0 | 0 | — |
case-05 | pass→pass | 24,931 | 24,038 | -4% | 1 | 1 | 0% | 3,660 | 5,054 | +38% | 0 | 0 | — |
case-06 | pass→pass | 26,398 | 26,676 | +1% | 1 | 1 | 0% | 3,837 | 5,289 | +38% | 0 | 0 | — |
case-07 | fail→pass | 15,062 | 4,002 | -73% | 1 | 1 | 0% | 2,319 | 2,066 | -11% | 0 | 0 | — |
case-08 | fail→pass | 21,783 | 22,027 | +1% | 1 | 1 | 0% | 3,166 | 4,624 | +46% | 0 | 0 | — |
case-09 | pass→pass | 22,560 | 20,302 | -10% | 1 | 1 | 0% | 3,279 | 4,489 | +37% | 0 | 0 | — |
case-10 | pass→pass | 21,167 | 21,433 | +1% | 1 | 1 | 0% | 2,980 | 4,469 | +50% | 0 | 0 | — |
case-11 | pass→pass | 19,609 | 21,710 | +11% | 1 | 1 | 0% | 2,839 | 4,658 | +64% | 0 | 0 | — |
case-12 | fail→pass | 17,869 | 16,372 | -8% | 1 | 1 | 0% | 2,756 | 4,112 | +49% | 0 | 0 | — |
case-13 | fail→pass | 22,623 | 22,009 | -3% | 1 | 1 | 0% | 3,429 | 4,780 | +39% | 0 | 0 | — |
case-14 | pass→pass | 14,342 | 15,905 | +11% | 1 | 1 | 0% | 2,357 | 4,076 | +73% | 0 | 0 | — |
case-15 | pass→pass | 23,287 | 20,141 | -14% | 1 | 1 | 0% | 3,433 | 4,488 | +31% | 0 | 0 | — |
case-16 | pass→pass | 21,986 | 21,301 | -3% | 1 | 1 | 0% | 3,184 | 4,519 | +42% | 0 | 0 | — |
case-17 | pass→pass | 19,290 | 19,534 | +1% | 1 | 1 | 0% | 2,915 | 4,368 | +50% | 0 | 0 | — |
case-18 | pass→pass | 18,081 | 16,777 | -7% | 1 | 1 | 0% | 2,986 | 4,117 | +38% | 0 | 0 | — |
case-19 | fail→pass | 7,792 | 1,969 | -75% | 1 | 1 | 0% | 1,234 | 1,706 | +38% | 0 | 0 | — |
case-20 | pass→pass | 18,864 | 17,271 | -8% | 1 | 1 | 0% | 2,916 | 3,962 | +36% | 0 | 0 | — |
case-21 | pass→pass | 17,090 | 18,703 | +9% | 1 | 1 | 0% | 2,580 | 4,153 | +61% | 0 | 0 | — |
case-22 | pass→pass | 19,262 | 17,761 | -8% | 1 | 1 | 0% | 2,814 | 4,164 | +48% | 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.