Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股可转债套利/转股溢价策略。当用户说"转债套利"、"可转债套利"、"convertible arb"、"转股溢价套利"、"转债对冲"、"转债Delta"时触发。基于 cn-stock-data 获取数据，分析可转债套利机会。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-convertible-arb/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-22 | ✓→✗ | ▼ Worse | -59% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 12% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 38% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 21% | 0% |
通过 cn-stock-data skill 获取数据：
# [转债] 套利分析报告
## 一、估值分析
| 指标 | 数值 |
|------|------|
| 转股溢价率 | 5.2% |
## 二、套利机会
[策略类型、预期收益]
## 三、对冲方案
[Delta对冲比例]
## 四、风险提示## [转债] 套利速览
- 转股溢价率 5.2%，中等水平
- 纯债价值 92，债底保护尚可
- 无折价套利机会
- 关注下修转股价可能性参考 references/convertible-arb-guide.md 获取详细方法论与 A股实证研究。
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-04 | pass→pass | 15,151 | 12,884 | -15% | 1 | 1 | 0% | 2,393 | 2,683 | +12% | 0 | 0 | — |
case-05 | fail→fail | 10,510 | 13,438 | +28% | 1 | 1 | 0% | 1,947 | 3,316 | +70% | 0 | 0 | — |
case-01 | pass→pass | 17,358 | 25,714 | +48% | 1 | 1 | 0% | 2,585 | 3,578 | +38% | 0 | 0 | — |
case-02 | fail→fail | 36,425 | 30,099 | -17% | 1 | 1 | 0% | 5,619 | 5,832 | +4% | 0 | 0 | — |
case-03 | pass→pass | 17,664 | 16,321 | -8% | 1 | 1 | 0% | 3,025 | 3,721 | +23% | 0 | 0 | — |
case-06 | pass→pass | 11,475 | 11,602 | +1% | 1 | 1 | 0% | 1,941 | 2,343 | +21% | 0 | 0 | — |
case-07 | pass→pass | 22,451 | 23,027 | +3% | 1 | 1 | 0% | 3,200 | 4,393 | +37% | 0 | 0 | — |
case-08 | pass→pass | 19,555 | 17,218 | -12% | 1 | 1 | 0% | 2,551 | 3,189 | +25% | 0 | 0 | — |
case-09 | pass→pass | 14,238 | 16,868 | +18% | 1 | 1 | 0% | 2,219 | 3,400 | +53% | 0 | 0 | — |
case-10 | pass→pass | 14,011 | 15,111 | +8% | 1 | 1 | 0% | 1,942 | 2,889 | +49% | 0 | 0 | — |
case-11 | pass→pass | 16,362 | 18,203 | +11% | 1 | 1 | 0% | 2,467 | 3,211 | +30% | 0 | 0 | — |
case-12 | pass→pass | 22,078 | 21,035 | -5% | 1 | 1 | 0% | 2,824 | 3,931 | +39% | 0 | 0 | — |
case-13 | pass→pass | 17,863 | 18,039 | +1% | 1 | 1 | 0% | 2,637 | 3,392 | +29% | 0 | 0 | — |
case-14 | pass→pass | 19,325 | 15,789 | -18% | 1 | 1 | 0% | 2,691 | 2,865 | +6% | 0 | 0 | — |
case-15 | pass→pass | 6,704 | 4,863 | -27% | 1 | 1 | 0% | 1,291 | 1,538 | +19% | 0 | 0 | — |
case-16 | pass→pass | 21,942 | 18,562 | -15% | 1 | 1 | 0% | 3,421 | 3,714 | +9% | 0 | 0 | — |
case-17 | pass→pass | 12,684 | 11,749 | -7% | 1 | 1 | 0% | 1,846 | 2,245 | +22% | 0 | 0 | — |
case-18 | pass→pass | 25,110 | 23,088 | -8% | 1 | 1 | 0% | 3,984 | 3,637 | -9% | 0 | 0 | — |
case-19 | fail→fail | 19,960 | 19,608 | -2% | 1 | 1 | 0% | 3,181 | 3,776 | +19% | 0 | 0 | — |
case-20 | pass→pass | 26,831 | 28,780 | +7% | 1 | 1 | 0% | 4,734 | 5,824 | +23% | 0 | 0 | — |
case-21 | pass→pass | 24,110 | 25,957 | +8% | 1 | 1 | 0% | 3,520 | 4,591 | +30% | 0 | 0 | — |
case-22 | pass→fail | 20,863 | 10,293 | -51% | 1 | 1 | 0% | 3,304 | 1,366 | -59% | 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, and 21 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 -5 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Other measured skills in the registry, with their headline benchmark lift.