Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股报价动态/最优报价分析。当用户说"报价动态"、"quote dynamics"、"最优报价"、"报价变化"、"挂单变化"、"报价行为"时触发。基于 cn-stock-data 获取数据，分析报价动态与最优报价策略。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-quote-dynamics/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-05 | ✓→✗ | ▼ Worse | -61% | 0% |
通过 cn-stock-data skill 获取数据：
# [标的] 报价动态分析报告
## 一、报价统计
| 指标 | 数值 |
|------|------|
| BBO变化频率 | 12次/分钟 |
## 二、信息含量
[报价vs成交的信息贡献]
## 三、报价模式
[日内模式、异常行为]
## 四、策略建议## [标的] 报价速览
- BBO变化 12次/分钟，活跃
- 报价持续时间 5.2秒
- 报价信息贡献 35%
- 报价行为正常，无异常参考 references/quote-dynamics-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 | 34,399 | 57,198 | +66% | 1 | 1 | 0% | 5,119 | 8,279 | +62% | 0 | 0 | — |
case-02 | fail→fail | 14,661 | 10,469 | -29% | 1 | 1 | 0% | 2,093 | 2,197 | +5% | 0 | 0 | — |
case-03 | fail→fail | 33,391 | 39,121 | +17% | 1 | 1 | 0% | 4,914 | 6,785 | +38% | 0 | 0 | — |
case-04 | pass→pass | 22,033 | 22,393 | +2% | 1 | 1 | 0% | 3,936 | 4,021 | +2% | 0 | 0 | — |
case-05 | pass→fail | 19,861 | 11,362 | -43% | 1 | 1 | 0% | 2,916 | 1,127 | -61% | 0 | 0 | — |
case-06 | pass→pass | 23,099 | 22,102 | -4% | 1 | 1 | 0% | 3,119 | 4,004 | +28% | 0 | 0 | — |
case-07 | fail→pass | 20,261 | 18,844 | -7% | 1 | 1 | 0% | 3,211 | 3,835 | +19% | 0 | 0 | — |
case-08 | pass→pass | 17,919 | 27,718 | +55% | 1 | 1 | 0% | 3,145 | 4,745 | +51% | 0 | 0 | — |
case-09 | pass→pass | 22,087 | 22,576 | +2% | 1 | 1 | 0% | 3,319 | 4,219 | +27% | 0 | 0 | — |
case-10 | pass→pass | 25,401 | 23,700 | -7% | 1 | 1 | 0% | 3,405 | 4,161 | +22% | 0 | 0 | — |
case-11 | fail→fail | 25,629 | 24,754 | -3% | 1 | 1 | 0% | 3,733 | 4,317 | +16% | 0 | 0 | — |
case-12 | pass→pass | 23,492 | 23,063 | -2% | 1 | 1 | 0% | 3,610 | 3,874 | +7% | 0 | 0 | — |
case-13 | pass→pass | 28,774 | 26,700 | -7% | 1 | 1 | 0% | 4,563 | 4,681 | +3% | 0 | 0 | — |
case-14 | fail→pass | 30,646 | 25,087 | -18% | 1 | 1 | 0% | 3,963 | 4,353 | +10% | 0 | 0 | — |
case-15 | pass→pass | 20,272 | 23,899 | +18% | 1 | 1 | 0% | 2,829 | 3,869 | +37% | 0 | 0 | — |
case-16 | fail→fail | 16,367 | 22,184 | +36% | 1 | 1 | 0% | 2,421 | 2,644 | +9% | 0 | 0 | — |
case-17 | fail→fail | 32,394 | 41,879 | +29% | 1 | 1 | 0% | 4,724 | 6,179 | +31% | 0 | 0 | — |
case-18 | pass→pass | 20,077 | 3,014 | -85% | 1 | 1 | 0% | 3,061 | 1,121 | -63% | 0 | 0 | — |
case-19 | fail→pass | 22,269 | 25,297 | +14% | 1 | 1 | 0% | 3,247 | 3,869 | +19% | 0 | 0 | — |
case-20 | fail→pass | 27,763 | 28,315 | +2% | 1 | 1 | 0% | 3,832 | 4,484 | +17% | 0 | 0 | — |
case-21 | fail→fail | 27,602 | 32,162 | +17% | 1 | 1 | 0% | 3,853 | 5,079 | +32% | 0 | 0 | — |
case-22 | pass→pass | 24,730 | 20,435 | -17% | 1 | 1 | 0% | 3,480 | 3,616 | +4% | 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 +14 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.
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.