Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股收益率分布/统计特征分析。当用户说"收益率分布"、"distribution"、"正态检验"、"偏度"、"峰度"、"收益率统计"时触发。量化分析收益率分布特征。支持formal和brief风格。
.claude/skills/aifinlab-a-share-distribution-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-24 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-06 | ✓→✗ | ▼ Worse | -82% | 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]
均值/中位数/标准差/偏度/峰度/最大值/最小值
拟合t分布/GED分布/混合正态，比较拟合优度
| 维度 | formal | brief | |------|--------|-------| | 统计量 | 完整描述性统计 | 偏度/峰度 | | 正态检验 | 多种检验结果 | 是否正态 | | 分布拟合 | 最佳拟合分布 | 分布类型 | 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 | 15,335 | 32,061 | +109% | 1 | 1 | 0% | 2,374 | 915 | -61% | 0 | 0 | — |
case-02 | fail→fail | 30,740 | 11,058 | -64% | 1 | 1 | 0% | 6,091 | 1,024 | -83% | 0 | 0 | — |
case-03 | fail→fail | 11,221 | 55,355 | +393% | 1 | 1 | 0% | 2,049 | 9,207 | +349% | 0 | 0 | — |
case-04 | fail→fail | 20,072 | 12,262 | -39% | 1 | 1 | 0% | 3,029 | 967 | -68% | 0 | 0 | — |
case-05 | fail→fail | 20,380 | 9,219 | -55% | 1 | 1 | 0% | 3,425 | 796 | -77% | 0 | 0 | — |
case-06 | pass→fail | 33,879 | 8,275 | -76% | 1 | 1 | 0% | 5,041 | 891 | -82% | 0 | 0 | — |
case-07 | pass→pass | 15,856 | 15,629 | -1% | 1 | 1 | 0% | 2,197 | 2,886 | +31% | 0 | 0 | — |
case-22 | pass→pass | 29,323 | 14,060 | -52% | 1 | 1 | 0% | 4,321 | 3,011 | -30% | 0 | 0 | — |
case-08 | pass→pass | 14,389 | 12,054 | -16% | 1 | 1 | 0% | 2,259 | 2,187 | -3% | 0 | 0 | — |
case-09 | pass→pass | 15,052 | 10,005 | -34% | 1 | 1 | 0% | 2,272 | 2,004 | -12% | 0 | 0 | — |
case-10 | fail→fail | 21,584 | 23,082 | +7% | 1 | 1 | 0% | 3,100 | 3,851 | +24% | 0 | 0 | — |
case-11 | pass→fail | 21,376 | 43,310 | +103% | 1 | 1 | 0% | 3,439 | 8,424 | +145% | 0 | 0 | — |
case-12 | fail→pass | 18,492 | 5,347 | -71% | 1 | 1 | 0% | 3,482 | 1,543 | -56% | 0 | 0 | — |
case-13 | fail→pass | 19,703 | 17,788 | -10% | 1 | 1 | 0% | 3,017 | 3,272 | +8% | 0 | 0 | — |
case-14 | fail→pass | 20,549 | 8,219 | -60% | 1 | 1 | 0% | 3,424 | 1,907 | -44% | 0 | 0 | — |
case-15 | pass→pass | 12,999 | 12,935 | -0% | 1 | 1 | 0% | 2,006 | 2,191 | +9% | 0 | 0 | — |
case-16 | pass→pass | 18,995 | 16,657 | -12% | 1 | 1 | 0% | 3,051 | 2,877 | -6% | 0 | 0 | — |
case-17 | fail→fail | 12,971 | 44,140 | +240% | 1 | 1 | 0% | 2,185 | 1,514 | -31% | 0 | 0 | — |
case-18 | pass→pass | 9,027 | 2,497 | -72% | 1 | 1 | 0% | 1,315 | 875 | -33% | 0 | 0 | — |
case-19 | pass→pass | 7,052 | 8,272 | +17% | 1 | 1 | 0% | 1,337 | 1,699 | +27% | 0 | 0 | — |
case-20 | pass→pass | 25,991 | 30,448 | +17% | 1 | 1 | 0% | 4,402 | 5,965 | +36% | 0 | 0 | — |
case-21 | pass→pass | 27,513 | 55,664 | +102% | 1 | 1 | 0% | 4,231 | 6,618 | +56% | 0 | 0 | — |
case-23 | pass→pass | 21,613 | 22,139 | +2% | 1 | 1 | 0% | 3,315 | 3,526 | +6% | 0 | 0 | — |
case-24 | fail→pass | 17,581 | 14,227 | -19% | 1 | 1 | 0% | 2,706 | 2,416 | -11% | 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. 24 cases were attempted, and 16 counted toward the lift figure. The other 8 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 +8 percentage points is the difference between those two pass rates over the 16 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.