Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股择时策略/大盘择时信号。当用户说"择时"、"market timing"、"该不该进场"、"现在能买吗"、"仓位建议"、"大盘择时"时触发。量化分析大盘择时信号。支持formal和brief风格。
.claude/skills/aifinlab-a-share-market-timing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-14 | ✓→✗ | ▼ Worse | -39% | 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]
各维度加权打分(0-100)
| 维度 | formal | brief | |------|--------|-------| | 各维度评分 | 详细指标明细 | 综合分数 | | 仓位建议 | 分档仓位表 | 建议仓位 | | 历史对照 | 类似评分历史表现 | 无 | 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 | 13,197 | 10,530 | -20% | 1 | 1 | 0% | 1,738 | 1,009 | -42% | 0 | 0 | — |
case-02 | fail→fail | 34,117 | 8,484 | -75% | 1 | 1 | 0% | 5,751 | 925 | -84% | 0 | 0 | — |
case-03 | fail→fail | 20,131 | 10,144 | -50% | 1 | 1 | 0% | 2,936 | 988 | -66% | 0 | 0 | — |
case-04 | fail→fail | 22,755 | 36,236 | +59% | 1 | 1 | 0% | 3,380 | 1,547 | -54% | 0 | 0 | — |
case-05 | pass→pass | 18,546 | 19,411 | +5% | 1 | 1 | 0% | 2,868 | 3,127 | +9% | 0 | 0 | — |
case-06 | pass→pass | 19,870 | 18,571 | -7% | 1 | 1 | 0% | 2,832 | 2,753 | -3% | 0 | 0 | — |
case-07 | pass→pass | 31,260 | 14,602 | -53% | 1 | 1 | 0% | 2,382 | 2,624 | +10% | 0 | 0 | — |
case-08 | fail→pass | 12,194 | 5,152 | -58% | 1 | 1 | 0% | 2,322 | 1,452 | -37% | 0 | 0 | — |
case-09 | fail→pass | 19,355 | 7,116 | -63% | 1 | 1 | 0% | 3,239 | 1,513 | -53% | 0 | 0 | — |
case-10 | fail→pass | 18,631 | 4,812 | -74% | 1 | 1 | 0% | 2,526 | 1,376 | -46% | 0 | 0 | — |
case-11 | pass→pass | 21,361 | 12,195 | -43% | 1 | 1 | 0% | 2,848 | 2,334 | -18% | 0 | 0 | — |
case-12 | pass→pass | 19,406 | 17,566 | -9% | 1 | 1 | 0% | 2,597 | 2,956 | +14% | 0 | 0 | — |
case-13 | pass→pass | 22,703 | 35,280 | +55% | 1 | 1 | 0% | 3,713 | 3,999 | +8% | 0 | 0 | — |
case-14 | pass→fail | 11,277 | 13,229 | +17% | 1 | 1 | 0% | 1,376 | 843 | -39% | 0 | 0 | — |
case-15 | fail→fail | 34,217 | 9,672 | -72% | 1 | 1 | 0% | 5,071 | 971 | -81% | 0 | 0 | — |
case-16 | pass→pass | 22,757 | 26,766 | +18% | 1 | 1 | 0% | 2,917 | 3,772 | +29% | 0 | 0 | — |
case-17 | pass→pass | 20,964 | 20,024 | -4% | 1 | 1 | 0% | 3,050 | 3,197 | +5% | 0 | 0 | — |
case-18 | fail→pass | 16,075 | 11,589 | -28% | 1 | 1 | 0% | 2,682 | 2,504 | -7% | 0 | 0 | — |
case-19 | pass→fail | 37,674 | 61,715 | +64% | 1 | 1 | 0% | 5,701 | 1,001 | -82% | 0 | 0 | — |
case-20 | pass→pass | 28,361 | 33,018 | +16% | 1 | 1 | 0% | 4,175 | 5,469 | +31% | 0 | 0 | — |
case-21 | pass→fail | 21,668 | 9,817 | -55% | 1 | 1 | 0% | 4,729 | 1,189 | -75% | 0 | 0 | — |
case-22 | pass→pass | 18,337 | 15,403 | -16% | 1 | 1 | 0% | 2,748 | 3,140 | +14% | 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 14 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 +5 percentage points is the difference between those two pass rates over the 14 comparable cases. 7 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.