Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股成交量时钟/交易节奏分析。当用户说"成交量时钟"、"volume clock"、"交易节奏"、"成交量分布"、"量能节奏"、"分时成交量"、"成交量模式"时触发。基于 cn-stock-data 获取数据，分析日内成交量分布、交易节奏异常、成交量预测。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-volume-clock/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 30% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-19 | ✓→✗ | ▼ Worse | -80% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 34% | 0% |
| case-08 | ✓→✓ | ＝ Same ✓ | 63% | 0% |
通过 cn-stock-data skill 获取数据：
# [标的] 成交量时钟分析报告
## 一、日内量分布
| 时段 | 占比 | 历史均值 | 偏离 |
|------|------|----------|------|
## 二、成交量时钟
[Volume Bar分析、信息密度]
## 三、节奏异常
[异常时段、可能原因]
## 四、量能预测
[全天量预估、尾盘流动性]## [标的] 量能节奏速览
- 当前成交占预估全天量的45%(时间过半)
- 10:15异常放量(偏离+120%)，大单涌入
- 预估全天成交额 8.5 亿
- 尾盘流动性预计充足参考 references/volume-clock-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-01 | fail→fail | 34,451 | 16,339 | -53% | 1 | 1 | 0% | 5,453 | 1,655 | -70% | 0 | 0 | — |
case-02 | fail→fail | 20,969 | 13,255 | -37% | 1 | 1 | 0% | 2,832 | 1,639 | -42% | 0 | 0 | — |
case-03 | fail→fail | 56,608 | 39,113 | -31% | 1 | 1 | 0% | 5,640 | 7,280 | +29% | 0 | 0 | — |
case-04 | fail→fail | 13,749 | 14,515 | +6% | 1 | 1 | 0% | 1,940 | 1,863 | -4% | 0 | 0 | — |
case-05 | pass→pass | 24,833 | 32,650 | +31% | 1 | 1 | 0% | 4,217 | 5,638 | +34% | 0 | 0 | — |
case-06 | fail→pass | 22,795 | 25,811 | +13% | 1 | 1 | 0% | 3,432 | 4,466 | +30% | 0 | 0 | — |
case-07 | fail→pass | 24,417 | 18,878 | -23% | 1 | 1 | 0% | 3,518 | 3,650 | +4% | 0 | 0 | — |
case-08 | pass→pass | 12,145 | 18,553 | +53% | 1 | 1 | 0% | 2,198 | 3,574 | +63% | 0 | 0 | — |
case-09 | pass→pass | 26,614 | 24,768 | -7% | 1 | 1 | 0% | 4,139 | 4,234 | +2% | 0 | 0 | — |
case-10 | fail→fail | 20,759 | 22,814 | +10% | 1 | 1 | 0% | 2,927 | 3,641 | +24% | 0 | 0 | — |
case-11 | pass→pass | 23,680 | 26,849 | +13% | 1 | 1 | 0% | 3,005 | 3,883 | +29% | 0 | 0 | — |
case-12 | fail→fail | 30,140 | 31,544 | +5% | 1 | 1 | 0% | 3,927 | 5,479 | +40% | 0 | 0 | — |
case-13 | fail→fail | 11,957 | 10,404 | -13% | 1 | 1 | 0% | 1,936 | 1,112 | -43% | 0 | 0 | — |
case-14 | pass→pass | 24,861 | 23,854 | -4% | 1 | 1 | 0% | 3,787 | 4,723 | +25% | 0 | 0 | — |
case-15 | pass→pass | 14,749 | 17,292 | +17% | 1 | 1 | 0% | 1,919 | 3,287 | +71% | 0 | 0 | — |
case-16 | pass→pass | 22,642 | 19,632 | -13% | 1 | 1 | 0% | 2,881 | 3,281 | +14% | 0 | 0 | — |
case-17 | fail→fail | 28,691 | 17,350 | -40% | 1 | 1 | 0% | 3,624 | 3,403 | -6% | 0 | 0 | — |
case-18 | pass→pass | 25,331 | 19,332 | -24% | 1 | 1 | 0% | 3,264 | 3,820 | +17% | 0 | 0 | — |
case-19 | pass→fail | 31,963 | 12,461 | -61% | 1 | 1 | 0% | 6,134 | 1,233 | -80% | 0 | 0 | — |
case-20 | pass→pass | 27,996 | 35,667 | +27% | 1 | 1 | 0% | 5,145 | 7,768 | +51% | 0 | 0 | — |
case-21 | pass→pass | 21,144 | 36,835 | +74% | 1 | 1 | 0% | 4,405 | 8,122 | +84% | 0 | 0 | — |
case-22 | pass→pass | 23,005 | 22,215 | -3% | 1 | 1 | 0% | 3,431 | 4,298 | +25% | 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 17 counted toward the lift figure. The other 5 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 17 comparable cases. 3 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.
Other measured skills in the registry, with their headline benchmark lift.