Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股协整检验/长期均衡关系分析。当用户说"协整"、"cointegration"、"长期均衡"、"价差平稳"、"XX和YY协整吗"时触发。量化检验股票间的协整关系。支持formal和brief风格。
.claude/skills/aifinlab-a-share-cointegration-test/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -72% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-19 | ✓→✗ | ▼ Worse | -74% | 0% |
bashSCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [日期] python "$SCRIPTS/cn_stock_data.py" quote --code [CODE] python "$SCRIPTS/cn_stock_data.py" finance --code [CODE]
估计长期均衡关系和短期调整速度
| 维度 | formal | brief | |------|--------|-------| | 单位根 | ADF+KPSS结果 | 是否I(1) | | 协整 | 检验统计量+p值 | 是否协整 | | 价差分析 | 半衰期+均值回归参数 | 配对可行性 | 默认风格：brief。
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 | 21,277 | 26,042 | +22% | 1 | 1 | 0% | 3,969 | 823 | -79% | 0 | 0 | — |
case-02 | fail→fail | 26,079 | 44,694 | +71% | 1 | 1 | 0% | 4,407 | 1,248 | -72% | 0 | 0 | — |
case-03 | fail→fail | 19,645 | 12,475 | -36% | 1 | 1 | 0% | 3,272 | 1,269 | -61% | 0 | 0 | — |
case-04 | fail→fail | 25,368 | 16,375 | -35% | 1 | 1 | 0% | 3,921 | 2,512 | -36% | 0 | 0 | — |
case-05 | pass→pass | 21,099 | 14,653 | -31% | 1 | 1 | 0% | 3,084 | 2,457 | -20% | 0 | 0 | — |
case-21 | pass→pass | 29,334 | 29,828 | +2% | 1 | 1 | 0% | 4,551 | 5,247 | +15% | 0 | 0 | — |
case-06 | pass→pass | 18,446 | 19,616 | +6% | 1 | 1 | 0% | 2,588 | 3,630 | +40% | 0 | 0 | — |
case-07 | pass→pass | 13,413 | 6,943 | -48% | 1 | 1 | 0% | 2,207 | 1,652 | -25% | 0 | 0 | — |
case-08 | fail→pass | 20,266 | 11,088 | -45% | 1 | 1 | 0% | 2,821 | 2,270 | -20% | 0 | 0 | — |
case-09 | pass→pass | 19,425 | 15,199 | -22% | 1 | 1 | 0% | 2,753 | 3,068 | +11% | 0 | 0 | — |
case-10 | pass→pass | 14,448 | 14,811 | +3% | 1 | 1 | 0% | 2,181 | 2,663 | +22% | 0 | 0 | — |
case-11 | fail→fail | 13,595 | 31,582 | +132% | 1 | 1 | 0% | 2,009 | 1,394 | -31% | 0 | 0 | — |
case-12 | pass→pass | 14,456 | 9,364 | -35% | 1 | 1 | 0% | 2,304 | 1,897 | -18% | 0 | 0 | — |
case-13 | fail→fail | 19,918 | 5,156 | -74% | 1 | 1 | 0% | 3,286 | 1,210 | -63% | 0 | 0 | — |
case-14 | fail→pass | 18,612 | 3,293 | -82% | 1 | 1 | 0% | 3,593 | 998 | -72% | 0 | 0 | — |
case-15 | pass→pass | 20,445 | 11,962 | -41% | 1 | 1 | 0% | 2,655 | 2,405 | -9% | 0 | 0 | — |
case-16 | fail→pass | 18,119 | 8,612 | -52% | 1 | 1 | 0% | 2,869 | 1,962 | -32% | 0 | 0 | — |
case-17 | pass→pass | 21,002 | 19,543 | -7% | 1 | 1 | 0% | 3,626 | 3,409 | -6% | 0 | 0 | — |
case-18 | fail→pass | 19,903 | 10,213 | -49% | 1 | 1 | 0% | 2,979 | 2,293 | -23% | 0 | 0 | — |
case-19 | pass→fail | 23,611 | 11,258 | -52% | 1 | 1 | 0% | 3,896 | 1,022 | -74% | 0 | 0 | — |
case-20 | pass→fail | 38,043 | 12,375 | -67% | 1 | 1 | 0% | 5,792 | 1,439 | -75% | 0 | 0 | — |
case-22 | pass→fail | 19,856 | 10,735 | -46% | 1 | 1 | 0% | 3,020 | 1,059 | -65% | 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 15 counted toward the lift figure. The other 7 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 15 comparable cases. 6 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/28/2026 | +5% |
Other measured skills in the registry, with their headline benchmark lift.