Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股委托不平衡/买卖盘口分析。当用户说"委托不平衡"、"order imbalance"、"买卖盘"、"盘口"、"挂单"、"买盘强还是卖盘强"时触发。量化分析盘口委托数据。支持formal和brief风格。
.claude/skills/aifinlab-a-share-order-imbalance/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 106% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -62% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -63% | 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]
OIB = (买委托量 - 卖委托量) / (买委托量 + 卖委托量)
| 维度 | formal | brief | |------|--------|-------| | OIB | 时序分析+统计 | 当前OIB | | 盘口 | 五档详细分析 | 买/卖主导 | 默认风格：brief。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,591 | 8,895 | -46% | 1 | 1 | 0% | 2,377 | 676 | -72% | 0 | 0 | — |
case-02 | fail→fail | 21,188 | 7,458 | -65% | 1 | 1 | 0% | 3,001 | 700 | -77% | 0 | 0 | — |
case-03 | fail→fail | 31,348 | 9,826 | -69% | 1 | 1 | 0% | 3,739 | 877 | -77% | 0 | 0 | — |
case-04 | pass→fail | 39,232 | 7,866 | -80% | 1 | 1 | 0% | 3,160 | 763 | -76% | 0 | 0 | — |
case-05 | fail→pass | 13,167 | 21,743 | +65% | 1 | 1 | 0% | 1,940 | 3,994 | +106% | 0 | 0 | — |
case-06 | pass→fail | 21,115 | 8,359 | -60% | 1 | 1 | 0% | 3,286 | 628 | -81% | 0 | 0 | — |
case-16 | fail→fail | 17,904 | 3,646 | -80% | 1 | 1 | 0% | 3,171 | 876 | -72% | 0 | 0 | — |
case-07 | fail→pass | 20,426 | 10,919 | -47% | 1 | 1 | 0% | 2,691 | 2,011 | -25% | 0 | 0 | — |
case-08 | fail→pass | 10,548 | 10,271 | -3% | 1 | 1 | 0% | 1,698 | 1,644 | -3% | 0 | 0 | — |
case-09 | pass→pass | 22,856 | 13,251 | -42% | 1 | 1 | 0% | 3,381 | 2,508 | -26% | 0 | 0 | — |
case-10 | pass→pass | 21,447 | 18,732 | -13% | 1 | 1 | 0% | 2,734 | 2,831 | +4% | 0 | 0 | — |
case-11 | pass→pass | 26,194 | 36,425 | +39% | 1 | 1 | 0% | 3,126 | 2,498 | -20% | 0 | 0 | — |
case-12 | fail→fail | 14,318 | 7,934 | -45% | 1 | 1 | 0% | 1,907 | 644 | -66% | 0 | 0 | — |
case-13 | fail→fail | 38,920 | 28,592 | -27% | 1 | 1 | 0% | 4,607 | 4,639 | +1% | 0 | 0 | — |
case-14 | pass→pass | 22,662 | 30,676 | +35% | 1 | 1 | 0% | 3,261 | 2,786 | -15% | 0 | 0 | — |
case-15 | pass→pass | 31,211 | 28,446 | -9% | 1 | 1 | 0% | 4,087 | 4,039 | -1% | 0 | 0 | — |
case-17 | fail→pass | 21,745 | 6,620 | -70% | 1 | 1 | 0% | 4,029 | 1,524 | -62% | 0 | 0 | — |
case-18 | pass→pass | 9,105 | 4,862 | -47% | 1 | 1 | 0% | 1,660 | 1,263 | -24% | 0 | 0 | — |
case-19 | fail→pass | 14,081 | 2,534 | -82% | 1 | 1 | 0% | 2,025 | 755 | -63% | 0 | 0 | — |
case-20 | pass→pass | 22,450 | 22,290 | -1% | 1 | 1 | 0% | 3,278 | 3,224 | -2% | 0 | 0 | — |
case-21 | fail→pass | 13,347 | 3,372 | -75% | 1 | 1 | 0% | 2,291 | 933 | -59% | 0 | 0 | — |
case-22 | pass→pass | 23,882 | 11,908 | -50% | 1 | 1 | 0% | 2,889 | 2,265 | -22% | 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. 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.