Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股交易分类/Lee-Ready算法。当用户说"交易分类"、"Lee-Ready"、"买卖分类"、"主动买卖"、"trade classification"、"BVC"时触发。基于 cn-stock-data 获取数据，对成交进行买卖方向分类。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-trade-classification/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-22 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-15 | ✓→✓ | ＝ Same ✓ | 2% | 0% |
通过 cn-stock-data skill 获取数据：
# [标的] 交易分类分析报告
## 一、分类结果
| 时段 | 主动买(万股) | 主动卖(万股) | 净买入 |
|------|------------|------------|--------|
## 二、分类方法
[使用的算法与准确率]
## 三、订单流分析
[OFI、净买入趋势]
## 四、信号解读## [标的] 交易分类速览
- 主动买入 58%，主动卖出 42%
- 净买入 +1,200万股
- 大单(>10万)净买入 +350万股
- 信号：资金偏多头参考 references/trade-classification-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-15 | pass→pass | 19,553 | 15,950 | -18% | 1 | 1 | 0% | 3,028 | 3,077 | +2% | 0 | 0 | — |
case-16 | pass→pass | 19,357 | 24,904 | +29% | 1 | 1 | 0% | 2,785 | 4,042 | +45% | 0 | 0 | — |
case-17 | pass→pass | 9,558 | 9,532 | -0% | 1 | 1 | 0% | 1,617 | 2,292 | +42% | 0 | 0 | — |
case-18 | fail→fail | 16,214 | 9,725 | -40% | 1 | 1 | 0% | 2,370 | 2,416 | +2% | 0 | 0 | — |
case-19 | pass→pass | 30,997 | 8,294 | -73% | 1 | 1 | 0% | 3,897 | 2,228 | -43% | 0 | 0 | — |
case-20 | pass→pass | 13,865 | 6,945 | -50% | 1 | 1 | 0% | 1,801 | 1,704 | -5% | 0 | 0 | — |
case-06 | fail→fail | 20,542 | 25,771 | +25% | 1 | 1 | 0% | 3,561 | 4,456 | +25% | 0 | 0 | — |
case-07 | pass→pass | 25,329 | 14,378 | -43% | 1 | 1 | 0% | 3,660 | 3,142 | -14% | 0 | 0 | — |
case-08 | pass→pass | 21,999 | 9,558 | -57% | 1 | 1 | 0% | 1,369 | 2,378 | +74% | 0 | 0 | — |
case-14 | fail→pass | 20,467 | 13,641 | -33% | 1 | 1 | 0% | 2,997 | 2,668 | -11% | 0 | 0 | — |
case-01 | fail→fail | 43,838 | 71,980 | +64% | 1 | 1 | 0% | 8,260 | 7,150 | -13% | 0 | 0 | — |
case-02 | fail→fail | 8,565 | 10,908 | +27% | 1 | 1 | 0% | 1,483 | 1,648 | +11% | 0 | 0 | — |
case-03 | fail→fail | 30,864 | 36,526 | +18% | 1 | 1 | 0% | 5,401 | 6,655 | +23% | 0 | 0 | — |
case-04 | fail→fail | 15,385 | 28,746 | +87% | 1 | 1 | 0% | 2,362 | 5,349 | +126% | 0 | 0 | — |
case-05 | fail→fail | 19,247 | 10,692 | -44% | 1 | 1 | 0% | 3,667 | 1,518 | -59% | 0 | 0 | — |
case-09 | pass→pass | 13,152 | 14,858 | +13% | 1 | 1 | 0% | 2,520 | 2,928 | +16% | 0 | 0 | — |
case-10 | pass→pass | 5,134 | 7,889 | +54% | 1 | 1 | 0% | 1,045 | 1,854 | +77% | 0 | 0 | — |
case-11 | pass→pass | 9,997 | 13,405 | +34% | 1 | 1 | 0% | 1,674 | 2,812 | +68% | 0 | 0 | — |
case-12 | pass→pass | 17,228 | 15,807 | -8% | 1 | 1 | 0% | 2,310 | 3,162 | +37% | 0 | 0 | — |
case-13 | fail→pass | 26,326 | 15,251 | -42% | 1 | 1 | 0% | 2,499 | 2,780 | +11% | 0 | 0 | — |
case-21 | fail→pass | 13,804 | 2,620 | -81% | 1 | 1 | 0% | 1,704 | 902 | -47% | 0 | 0 | — |
case-22 | fail→pass | 9,101 | 2,589 | -72% | 1 | 1 | 0% | 1,405 | 976 | -31% | 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 20 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 +18 percentage points is the difference between those two pass rates over the 20 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.