Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股价格发现/信息效率分析。当用户说"价格发现"、"price discovery"、"信息效率"、"定价效率"、"信息融入"、"价格领先"时触发。基于 cn-stock-data 获取数据，分析价格发现效率与信息融入速度。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-price-discovery/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-19 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-06 | ✓→✗ | ▼ Worse | 110% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 59% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 107% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | 21% | 0% |
通过 cn-stock-data skill 获取数据：
# [标的] 价格发现分析报告
## 一、信息效率
| 指标 | 数值 | 评估 |
|------|------|------|
## 二、跨市场分析
[A-H溢价、信息份额]
## 三、事件分析
[信息融入速度]
## 四、效率评估## [标的] 价格发现速览
- 方差比VR(5)=0.92，接近有效
- A-H溢价 +25%，A股偏贵
- 信息融入：公告后约30分钟完全反映
- 效率评级：中等(分析师覆盖不足)参考 references/price-discovery-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-21 | fail→fail | 10,050 | 7,630 | -24% | 1 | 1 | 0% | 1,456 | 1,707 | +17% | 0 | 0 | — |
case-01 | fail→fail | 44,912 | 37,881 | -16% | 1 | 1 | 0% | 7,111 | 5,892 | -17% | 0 | 0 | — |
case-02 | fail→fail | 15,977 | 8,356 | -48% | 1 | 1 | 0% | 2,556 | 2,012 | -21% | 0 | 0 | — |
case-03 | fail→fail | 43,127 | 33,620 | -22% | 1 | 1 | 0% | 6,875 | 6,142 | -11% | 0 | 0 | — |
case-04 | pass→pass | 34,552 | 40,655 | +18% | 1 | 1 | 0% | 5,604 | 8,895 | +59% | 0 | 0 | — |
case-05 | pass→pass | 23,111 | 53,360 | +131% | 1 | 1 | 0% | 3,888 | 8,060 | +107% | 0 | 0 | — |
case-06 | pass→fail | 25,729 | 53,204 | +107% | 1 | 1 | 0% | 4,223 | 8,882 | +110% | 0 | 0 | — |
case-07 | pass→pass | 21,038 | 25,040 | +19% | 1 | 1 | 0% | 3,653 | 4,415 | +21% | 0 | 0 | — |
case-08 | pass→pass | 20,477 | 20,608 | +1% | 1 | 1 | 0% | 3,329 | 3,902 | +17% | 0 | 0 | — |
case-09 | fail→fail | 12,156 | 19,097 | +57% | 1 | 1 | 0% | 1,998 | 2,930 | +47% | 0 | 0 | — |
case-10 | fail→fail | 21,228 | 19,607 | -8% | 1 | 1 | 0% | 2,780 | 3,452 | +24% | 0 | 0 | — |
case-11 | pass→pass | 22,389 | 24,061 | +7% | 1 | 1 | 0% | 3,417 | 4,463 | +31% | 0 | 0 | — |
case-12 | pass→pass | 26,460 | 27,863 | +5% | 1 | 1 | 0% | 3,158 | 4,141 | +31% | 0 | 0 | — |
case-13 | fail→fail | 39,690 | 34,633 | -13% | 1 | 1 | 0% | 5,919 | 5,224 | -12% | 0 | 0 | — |
case-14 | fail→fail | 15,917 | 13,210 | -17% | 1 | 1 | 0% | 2,517 | 2,301 | -9% | 0 | 0 | — |
case-15 | pass→pass | 15,448 | 18,548 | +20% | 1 | 1 | 0% | 2,443 | 3,587 | +47% | 0 | 0 | — |
case-16 | pass→pass | 28,416 | 23,365 | -18% | 1 | 1 | 0% | 4,146 | 4,084 | -1% | 0 | 0 | — |
case-17 | pass→pass | 31,718 | 32,161 | +1% | 1 | 1 | 0% | 4,018 | 5,762 | +43% | 0 | 0 | — |
case-18 | fail→fail | 41,785 | 56,233 | +35% | 1 | 1 | 0% | 6,449 | 7,297 | +13% | 0 | 0 | — |
case-19 | fail→pass | 20,722 | 25,460 | +23% | 1 | 1 | 0% | 2,537 | 3,675 | +45% | 0 | 0 | — |
case-20 | pass→pass | 14,382 | 15,746 | +9% | 1 | 1 | 0% | 2,177 | 3,064 | +41% | 0 | 0 | — |
case-22 | fail→fail | 35,082 | 43,656 | +24% | 1 | 1 | 0% | 5,542 | 5,676 | +2% | 0 | 0 | — |
case-23 | pass→pass | 30,969 | 25,910 | -16% | 1 | 1 | 0% | 4,055 | 4,996 | +23% | 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. 23 cases were attempted. The headline lift of 0 percentage points is the difference between those two pass rates over the 23 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/28/2026 | 0% |
Other measured skills in the registry, with their headline benchmark lift.