Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →当用户需要在银行对公金融场景下，对房地产企业或项目授信做项目去化、受限资金、债务到期压力、销售回款和政策合规约束判断时使用本技能。适合服务客户经理和地产审查岗。
.claude/skills/aifinlab-bank-t128-corporate-finance-credit-real-estate-assistant/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 58% | 0% |
这个 skill 面向房地产授信，不把“地产背景”当成抽象标签，而是按项目、资金和债务三个层面拆开判断。重点看项目能否卖、钱能否回、债能否还，以及政策和手续是否让这个项目具备继续推进的基础。
详细字段见 input-schema.md 和 output-schema.md。
scripts/run_skill.py：输出房地产授信判断包shared/corporate_credit_skill_engine.py：共享分析引擎，内含地产专项规则| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→fail | 40,518 | 35,777 | -12% | 1 | 1 | 0% | 3,812 | 5,999 | +57% | 0 | 0 | — |
case-02 | pass→pass | 10,212 | 20,279 | +99% | 1 | 1 | 0% | 1,422 | 3,996 | +181% | 0 | 0 | — |
case-03 | pass→pass | 21,000 | 21,472 | +2% | 1 | 1 | 0% | 2,714 | 4,560 | +68% | 0 | 0 | — |
case-04 | fail→pass | 15,309 | 15,498 | +1% | 1 | 1 | 0% | 2,289 | 3,845 | +68% | 0 | 0 | — |
case-05 | pass→pass | 20,193 | 21,981 | +9% | 1 | 1 | 0% | 2,670 | 4,199 | +57% | 0 | 0 | — |
case-06 | pass→pass | 20,595 | 24,743 | +20% | 1 | 1 | 0% | 2,997 | 4,565 | +52% | 0 | 0 | — |
case-07 | pass→pass | 22,946 | 28,571 | +25% | 1 | 1 | 0% | 2,762 | 4,976 | +80% | 0 | 0 | — |
case-08 | fail→pass | 20,361 | 24,545 | +21% | 1 | 1 | 0% | 2,976 | 4,723 | +59% | 0 | 0 | — |
case-09 | pass→pass | 16,204 | 22,964 | +42% | 1 | 1 | 0% | 2,417 | 4,433 | +83% | 0 | 0 | — |
case-10 | pass→pass | 25,419 | 22,627 | -11% | 1 | 1 | 0% | 3,653 | 4,811 | +32% | 0 | 0 | — |
case-11 | pass→pass | 20,667 | 30,878 | +49% | 1 | 1 | 0% | 2,526 | 4,057 | +61% | 0 | 0 | — |
case-12 | fail→pass | 30,706 | 19,530 | -36% | 1 | 1 | 0% | 3,095 | 4,308 | +39% | 0 | 0 | — |
case-13 | fail→pass | 20,704 | 22,553 | +9% | 1 | 1 | 0% | 2,707 | 4,585 | +69% | 0 | 0 | — |
case-14 | pass→pass | 26,521 | 22,718 | -14% | 1 | 1 | 0% | 3,333 | 4,570 | +37% | 0 | 0 | — |
case-15 | pass→pass | 22,764 | 25,344 | +11% | 1 | 1 | 0% | 2,894 | 4,413 | +52% | 0 | 0 | — |
case-16 | fail→pass | 25,610 | 28,032 | +9% | 1 | 1 | 0% | 3,154 | 4,990 | +58% | 0 | 0 | — |
case-17 | pass→pass | 23,288 | 33,354 | +43% | 1 | 1 | 0% | 2,928 | 6,017 | +105% | 0 | 0 | — |
case-18 | pass→pass | 25,159 | 28,272 | +12% | 1 | 1 | 0% | 3,323 | 5,354 | +61% | 0 | 0 | — |
case-19 | pass→pass | 15,009 | 21,535 | +43% | 1 | 1 | 0% | 2,124 | 4,059 | +91% | 0 | 0 | — |
case-20 | pass→pass | 24,798 | 22,419 | -10% | 1 | 1 | 0% | 3,042 | 4,660 | +53% | 0 | 0 | — |
case-21 | pass→pass | 24,027 | 17,365 | -28% | 1 | 1 | 0% | 3,168 | 4,271 | +35% | 0 | 0 | — |
case-22 | pass→pass | 22,841 | 19,886 | -13% | 1 | 1 | 0% | 2,911 | 4,330 | +49% | 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 +18 percentage points is the difference between those two pass rates over the 22 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.