Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股Beta对冲/市场中性策略。当用户说"对冲"、"beta hedging"、"市场中性"、"对冲策略"、"怎么对冲"、"空头对冲"时触发。量化构建市场中性组合。支持formal和brief风格。
.claude/skills/aifinlab-a-share-beta-hedging/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-24 | ✗→✓ | ▲ Improved | 16% | 0% |
bashSCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [日期] python "$SCRIPTS/cn_stock_data.py" quote --code [CODE] python "$SCRIPTS/cn_stock_data.py" finance --code [CODE]
| 维度 | formal | brief | |------|--------|-------| | Beta计算 | 多种方法对比 | 当前Beta | | 对冲方案 | 完整对冲方案 | 推荐工具+比率 | | 成本分析 | 基差+展期成本 | 年化对冲成本 | 默认风格：brief。
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-01 | fail→pass | 28,436 | 26,757 | -6% | 1 | 1 | 0% | 4,664 | 3,710 | -20% | 0 | 0 | — |
case-02 | fail→fail | 20,310 | 11,605 | -43% | 1 | 1 | 0% | 3,291 | 1,186 | -64% | 0 | 0 | — |
case-03 | fail→pass | 22,109 | 15,694 | -29% | 1 | 1 | 0% | 3,539 | 2,938 | -17% | 0 | 0 | — |
case-04 | pass→fail | 30,968 | 9,295 | -70% | 1 | 1 | 0% | 6,097 | 1,193 | -80% | 0 | 0 | — |
case-05 | pass→pass | 23,941 | 19,578 | -18% | 1 | 1 | 0% | 3,744 | 3,648 | -3% | 0 | 0 | — |
case-06 | pass→pass | 20,889 | 22,223 | +6% | 1 | 1 | 0% | 3,609 | 4,380 | +21% | 0 | 0 | — |
case-07 | pass→pass | 39,802 | 41,961 | +5% | 1 | 1 | 0% | 5,738 | 7,034 | +23% | 0 | 0 | — |
case-08 | pass→pass | 23,856 | 23,478 | -2% | 1 | 1 | 0% | 4,074 | 4,040 | -1% | 0 | 0 | — |
case-09 | fail→fail | 19,764 | 12,118 | -39% | 1 | 1 | 0% | 3,474 | 2,889 | -17% | 0 | 0 | — |
case-10 | pass→pass | 18,434 | 17,296 | -6% | 1 | 1 | 0% | 2,531 | 3,237 | +28% | 0 | 0 | — |
case-11 | pass→pass | 9,904 | 8,564 | -14% | 1 | 1 | 0% | 1,556 | 2,114 | +36% | 0 | 0 | — |
case-12 | pass→pass | 11,264 | 3,526 | -69% | 1 | 1 | 0% | 1,708 | 1,117 | -35% | 0 | 0 | — |
case-13 | pass→pass | 23,408 | 20,193 | -14% | 1 | 1 | 0% | 3,451 | 3,569 | +3% | 0 | 0 | — |
case-14 | pass→pass | 18,100 | 17,946 | -1% | 1 | 1 | 0% | 2,987 | 3,347 | +12% | 0 | 0 | — |
case-15 | fail→pass | 24,930 | 3,693 | -85% | 1 | 1 | 0% | 1,812 | 1,242 | -31% | 0 | 0 | — |
case-16 | fail→pass | 11,115 | 3,372 | -70% | 1 | 1 | 0% | 1,884 | 1,133 | -40% | 0 | 0 | — |
case-17 | pass→pass | 15,494 | 14,091 | -9% | 1 | 1 | 0% | 2,461 | 2,540 | +3% | 0 | 0 | — |
case-18 | pass→pass | 19,428 | 14,197 | -27% | 1 | 1 | 0% | 3,238 | 2,878 | -11% | 0 | 0 | — |
case-19 | pass→pass | 18,678 | 19,810 | +6% | 1 | 1 | 0% | 2,579 | 2,466 | -4% | 0 | 0 | — |
case-20 | pass→pass | 5,948 | 5,970 | +0% | 1 | 1 | 0% | 1,098 | 1,583 | +44% | 0 | 0 | — |
case-21 | pass→pass | 30,197 | 15,021 | -50% | 1 | 1 | 0% | 3,148 | 2,951 | -6% | 0 | 0 | — |
case-22 | pass→pass | 22,866 | 18,710 | -18% | 1 | 1 | 0% | 3,621 | 3,339 | -8% | 0 | 0 | — |
case-23 | pass→pass | 11,088 | 5,388 | -51% | 1 | 1 | 0% | 2,120 | 1,462 | -31% | 0 | 0 | — |
case-24 | fail→pass | 13,597 | 10,110 | -26% | 1 | 1 | 0% | 1,927 | 2,233 | +16% | 0 | 0 | — |
case-25 | pass→pass | 21,456 | 16,524 | -23% | 1 | 1 | 0% | 2,992 | 2,784 | -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. 25 cases were attempted, and 23 counted toward the lift figure. The other 2 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 +16 percentage points is the difference between those two pass rates over the 23 comparable cases. 2 cases got worse with the skill loaded, and they are 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/28/2026 | +14% |
Other measured skills in the registry, with their headline benchmark lift.