Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股跳空缺口分析/缺口策略。当用户说"缺口"、"gap"、"跳空"、"缺口分析"、"缺口回补"、"跳空高开"、"跳空低开"时触发。量化分析跳空缺口特征和交易含义。支持formal和brief风格。
.claude/skills/aifinlab-a-share-gap-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -28% | 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]
历史上同类缺口的回补概率和回补时间
| 维度 | formal | brief | |------|--------|-------| | 缺口列表 | 历史缺口全表 | 近期缺口 | | 分类 | 缺口类型判断 | 类型+含义 | | 回补统计 | 回补概率+时间 | 是否回补 | 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 | 23,892 | 9,286 | -61% | 1 | 1 | 0% | 3,707 | 948 | -74% | 0 | 0 | — |
case-02 | fail→fail | 27,466 | 11,771 | -57% | 1 | 1 | 0% | 4,728 | 1,159 | -75% | 0 | 0 | — |
case-03 | fail→fail | 21,570 | 9,616 | -55% | 1 | 1 | 0% | 2,499 | 936 | -63% | 0 | 0 | — |
case-04 | pass→pass | 15,627 | 23,445 | +50% | 1 | 1 | 0% | 2,231 | 2,983 | +34% | 0 | 0 | — |
case-05 | pass→pass | 17,571 | 20,480 | +17% | 1 | 1 | 0% | 2,290 | 3,257 | +42% | 0 | 0 | — |
case-06 | pass→pass | 12,221 | 14,182 | +16% | 1 | 1 | 0% | 1,795 | 2,652 | +48% | 0 | 0 | — |
case-16 | pass→pass | 13,394 | 12,822 | -4% | 1 | 1 | 0% | 1,996 | 2,565 | +29% | 0 | 0 | — |
case-17 | pass→pass | 12,089 | 8,859 | -27% | 1 | 1 | 0% | 2,194 | 2,037 | -7% | 0 | 0 | — |
case-07 | pass→pass | 14,968 | 18,696 | +25% | 1 | 1 | 0% | 2,137 | 2,887 | +35% | 0 | 0 | — |
case-08 | pass→pass | 24,598 | 22,510 | -8% | 1 | 1 | 0% | 3,197 | 3,646 | +14% | 0 | 0 | — |
case-09 | pass→pass | 20,447 | 37,801 | +85% | 1 | 1 | 0% | 2,665 | 3,231 | +21% | 0 | 0 | — |
case-10 | fail→pass | 16,288 | 11,201 | -31% | 1 | 1 | 0% | 2,375 | 1,937 | -18% | 0 | 0 | — |
case-11 | fail→pass | 17,245 | 6,390 | -63% | 1 | 1 | 0% | 3,209 | 1,547 | -52% | 0 | 0 | — |
case-12 | fail→pass | 17,735 | 5,488 | -69% | 1 | 1 | 0% | 3,010 | 1,445 | -52% | 0 | 0 | — |
case-13 | fail→pass | 14,017 | 18,957 | +35% | 1 | 1 | 0% | 2,044 | 2,754 | +35% | 0 | 0 | — |
case-14 | fail→pass | 17,171 | 14,809 | -14% | 1 | 1 | 0% | 2,604 | 1,873 | -28% | 0 | 0 | — |
case-15 | fail→pass | 12,085 | 1,823 | -85% | 1 | 1 | 0% | 1,652 | 705 | -57% | 0 | 0 | — |
case-18 | pass→pass | 11,748 | 9,029 | -23% | 1 | 1 | 0% | 1,633 | 1,744 | +7% | 0 | 0 | — |
case-19 | fail→pass | 14,617 | 17,162 | +17% | 1 | 1 | 0% | 1,918 | 2,560 | +33% | 0 | 0 | — |
case-20 | pass→fail | 25,417 | 7,681 | -70% | 1 | 1 | 0% | 3,925 | 897 | -77% | 0 | 0 | — |
case-21 | pass→fail | 18,502 | 11,354 | -39% | 1 | 1 | 0% | 2,736 | 1,206 | -56% | 0 | 0 | — |
case-22 | pass→fail | 22,150 | 11,815 | -47% | 1 | 1 | 0% | 3,317 | 976 | -71% | 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 16 counted toward the lift figure. The other 6 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 +18 percentage points is the difference between those two pass rates over the 16 comparable cases. 4 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.