Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股滑点分析/执行成本分析。当用户说"滑点"、"slippage"、"执行成本"、"成交偏差"、"滑点分析"、"实际成交价偏差"时触发。基于 cn-stock-data 获取数据，分析交易滑点与执行成本。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-slippage-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 2% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 16% | 0% |
通过 cn-stock-data skill 获取数据：
# 滑点分析报告
## 一、滑点统计
| 指标 | 数值 |
|------|------|
| 平均滑点 | 0.08% |
## 二、滑点分解
[冲击/延迟/价差各占比]
## 三、滑点模式
[时段/订单大小的影响]
## 四、优化建议## 滑点分析速览
- 平均滑点 0.08%
- 主要来源：市场冲击(60%)+价差(30%)
- 开盘时段滑点最大(0.15%)
- 建议：避开开盘15分钟，使用限价单参考 references/slippage-analysis-guide.md 获取详细方法论与 A股实证研究。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 28,043 | 46,556 | +66% | 1 | 1 | 0% | 4,474 | 6,480 | +45% | 0 | 0 | — |
case-02 | fail→fail | 12,066 | 12,639 | +5% | 1 | 1 | 0% | 1,969 | 2,641 | +34% | 0 | 0 | — |
case-03 | fail→fail | 31,944 | 25,716 | -19% | 1 | 1 | 0% | 4,434 | 4,786 | +8% | 0 | 0 | — |
case-04 | pass→pass | 23,544 | 19,463 | -17% | 1 | 1 | 0% | 3,922 | 3,992 | +2% | 0 | 0 | — |
case-05 | pass→pass | 23,837 | 29,579 | +24% | 1 | 1 | 0% | 4,173 | 4,841 | +16% | 0 | 0 | — |
case-06 | pass→pass | 24,742 | 25,123 | +2% | 1 | 1 | 0% | 3,604 | 4,310 | +20% | 0 | 0 | — |
case-07 | pass→pass | 11,726 | 13,417 | +14% | 1 | 1 | 0% | 2,238 | 2,653 | +19% | 0 | 0 | — |
case-08 | pass→pass | 12,441 | 16,238 | +31% | 1 | 1 | 0% | 2,356 | 3,001 | +27% | 0 | 0 | — |
case-09 | pass→pass | 21,435 | 19,549 | -9% | 1 | 1 | 0% | 3,156 | 3,125 | -1% | 0 | 0 | — |
case-10 | fail→fail | 12,947 | 15,058 | +16% | 1 | 1 | 0% | 1,963 | 2,978 | +52% | 0 | 0 | — |
case-11 | pass→pass | 23,053 | 21,474 | -7% | 1 | 1 | 0% | 3,011 | 3,791 | +26% | 0 | 0 | — |
case-12 | pass→pass | 22,570 | 19,192 | -15% | 1 | 1 | 0% | 3,375 | 3,581 | +6% | 0 | 0 | — |
case-13 | fail→fail | 23,192 | 30,077 | +30% | 1 | 1 | 0% | 3,456 | 5,047 | +46% | 0 | 0 | — |
case-14 | fail→pass | 19,126 | 8,566 | -55% | 1 | 1 | 0% | 2,915 | 1,775 | -39% | 0 | 0 | — |
case-15 | pass→pass | 21,757 | 17,034 | -22% | 1 | 1 | 0% | 3,054 | 3,082 | +1% | 0 | 0 | — |
case-16 | fail→pass | 18,138 | 20,099 | +11% | 1 | 1 | 0% | 2,930 | 3,249 | +11% | 0 | 0 | — |
case-17 | pass→pass | 23,121 | 18,181 | -21% | 1 | 1 | 0% | 3,429 | 3,438 | +0% | 0 | 0 | — |
case-18 | pass→pass | 16,919 | 16,223 | -4% | 1 | 1 | 0% | 2,181 | 2,470 | +13% | 0 | 0 | — |
case-19 | fail→fail | 22,597 | 15,069 | -33% | 1 | 1 | 0% | 3,470 | 3,451 | -1% | 0 | 0 | — |
case-20 | fail→pass | 20,288 | 16,644 | -18% | 1 | 1 | 0% | 3,075 | 3,132 | +2% | 0 | 0 | — |
case-21 | pass→pass | 21,136 | 18,921 | -10% | 1 | 1 | 0% | 3,116 | 3,320 | +7% | 0 | 0 | — |
case-22 | pass→pass | 25,399 | 21,688 | -15% | 1 | 1 | 0% | 3,430 | 4,062 | +18% | 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 +14 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.