Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →用于信托领域财富与家族信托中的受益安排解释助手场景，支持结构化处理与报告输出。
.claude/skills/aifinlab-benefit-arrangement-explainer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -49% | 0% |
本技能用于信托业务场景的结构化处理，支持数据解析、规则识别与报告输出。
id, name, text, status, timestamp，以及场景相关业务字段bashpython scripts/analyze.py --input input.json --output report.md
python# 调用 skill result = run_skill({ "param1": "value1", "param2": "value2" })
bashpython scripts/run_skill.py --input data.json
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | pass→pass | 15,914 | 7,211 | -55% | 1 | 1 | 0% | 2,117 | 1,504 | -29% | 0 | 0 | — |
case-11 | fail→fail | 9,275 | 2,763 | -70% | 1 | 1 | 0% | 1,345 | 812 | -40% | 0 | 0 | — |
case-01 | fail→pass | 54,873 | 48,989 | -11% | 1 | 1 | 0% | 3,857 | 3,698 | -4% | 0 | 0 | — |
case-02 | fail→pass | 9,778 | 33,464 | +242% | 1 | 1 | 0% | 1,640 | 1,006 | -39% | 0 | 0 | — |
case-03 | pass→pass | 38,968 | 32,944 | -15% | 1 | 1 | 0% | 1,586 | 821 | -48% | 0 | 0 | — |
case-04 | fail→pass | 15,255 | 15,155 | -1% | 1 | 1 | 0% | 2,478 | 2,689 | +9% | 0 | 0 | — |
case-05 | fail→fail | 8,212 | 33,459 | +307% | 1 | 1 | 0% | 1,186 | 932 | -21% | 0 | 0 | — |
case-07 | pass→pass | 11,641 | 6,747 | -42% | 1 | 1 | 0% | 1,712 | 1,378 | -20% | 0 | 0 | — |
case-08 | pass→pass | 16,242 | 4,897 | -70% | 1 | 1 | 0% | 2,231 | 1,131 | -49% | 0 | 0 | — |
case-09 | pass→pass | 7,170 | 2,172 | -70% | 1 | 1 | 0% | 1,155 | 722 | -37% | 0 | 0 | — |
case-10 | fail→pass | 10,974 | 3,776 | -66% | 1 | 1 | 0% | 1,615 | 909 | -44% | 0 | 0 | — |
case-12 | fail→pass | 10,310 | 3,172 | -69% | 1 | 1 | 0% | 1,643 | 846 | -49% | 0 | 0 | — |
case-13 | fail→pass | 9,453 | 1,863 | -80% | 1 | 1 | 0% | 1,491 | 647 | -57% | 0 | 0 | — |
case-14 | fail→pass | 9,419 | 2,931 | -69% | 1 | 1 | 0% | 1,334 | 812 | -39% | 0 | 0 | — |
case-15 | fail→pass | 10,137 | 2,639 | -74% | 1 | 1 | 0% | 1,767 | 805 | -54% | 0 | 0 | — |
case-16 | fail→pass | 22,619 | 2,258 | -90% | 1 | 1 | 0% | 1,017 | 703 | -31% | 0 | 0 | — |
case-17 | pass→pass | 12,599 | 14,044 | +11% | 1 | 1 | 0% | 1,882 | 2,110 | +12% | 0 | 0 | — |
case-18 | pass→fail | 8,229 | 2,300 | -72% | 1 | 1 | 0% | 1,200 | 674 | -44% | 0 | 0 | — |
case-19 | pass→pass | 11,664 | 8,603 | -26% | 1 | 1 | 0% | 1,786 | 1,711 | -4% | 0 | 0 | — |
case-20 | fail→fail | 6,180 | 3,023 | -51% | 1 | 1 | 0% | 940 | 873 | -7% | 0 | 0 | — |
case-21 | fail→fail | 16,549 | 20,945 | +27% | 1 | 1 | 0% | 2,622 | 3,316 | +26% | 0 | 0 | — |
case-22 | fail→pass | 13,088 | 6,484 | -50% | 1 | 1 | 0% | 1,468 | 1,325 | -10% | 0 | 0 | — |
case-23 | fail→fail | 18,700 | 16,425 | -12% | 1 | 1 | 0% | 3,244 | 3,018 | -7% | 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, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +39 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.