Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股突破策略/形态突破量化分析。当用户说"突破策略"、"breakout"、"形态突破"、"箱体突破"、"平台突破"、"新高突破"、"假突破"、"真突破"时触发。基于 cn-stock-data 获取K线数据，量化识别价格突破信号，评估突破有效性。支持研报风格（formal）和快速分析风格（brief）。
.claude/skills/aifinlab-a-share-breakout-strategy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-11 | ✓→✗ | ▼ Worse | -13% | 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]
获取至少 120 个交易日的 OHLCV 数据用于识别整理形态。
| 维度 | formal | brief | |------|--------|-------| | 形态识别 | 详细形态参数+图示描述 | 当前形态类型 | | 突破信号 | 量价配合度评分 | 有/无突破 | | 历史回测 | 突破后 5/10/20 日收益率 | 突破成功率 |
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 | 14,724 | 11,671 | -21% | 1 | 1 | 0% | 1,904 | 1,106 | -42% | 0 | 0 | — |
case-02 | fail→fail | 27,141 | 10,465 | -61% | 1 | 1 | 0% | 3,226 | 1,041 | -68% | 0 | 0 | — |
case-03 | pass→pass | 20,116 | 12,604 | -37% | 1 | 1 | 0% | 3,008 | 2,259 | -25% | 0 | 0 | — |
case-04 | fail→fail | 20,833 | 10,357 | -50% | 1 | 1 | 0% | 2,907 | 989 | -66% | 0 | 0 | — |
case-05 | pass→pass | 33,040 | 13,277 | -60% | 1 | 1 | 0% | 2,366 | 2,493 | +5% | 0 | 0 | — |
case-06 | pass→pass | 20,737 | 17,688 | -15% | 1 | 1 | 0% | 3,176 | 3,335 | +5% | 0 | 0 | — |
case-07 | pass→pass | 22,318 | 25,785 | +16% | 1 | 1 | 0% | 3,062 | 4,315 | +41% | 0 | 0 | — |
case-08 | pass→pass | 18,313 | 15,638 | -15% | 1 | 1 | 0% | 2,512 | 2,796 | +11% | 0 | 0 | — |
case-09 | pass→pass | 26,495 | 21,031 | -21% | 1 | 1 | 0% | 3,323 | 3,657 | +10% | 0 | 0 | — |
case-10 | pass→pass | 17,952 | 14,125 | -21% | 1 | 1 | 0% | 2,141 | 2,594 | +21% | 0 | 0 | — |
case-11 | pass→fail | 12,317 | 7,562 | -39% | 1 | 1 | 0% | 1,812 | 1,579 | -13% | 0 | 0 | — |
case-12 | pass→pass | 10,169 | 5,900 | -42% | 1 | 1 | 0% | 1,405 | 1,489 | +6% | 0 | 0 | — |
case-13 | fail→pass | 11,871 | 3,821 | -68% | 1 | 1 | 0% | 1,262 | 1,049 | -17% | 0 | 0 | — |
case-14 | pass→pass | 6,977 | 7,649 | +10% | 1 | 1 | 0% | 1,254 | 1,241 | -1% | 0 | 0 | — |
case-15 | pass→pass | 17,843 | 12,522 | -30% | 1 | 1 | 0% | 2,417 | 2,747 | +14% | 0 | 0 | — |
case-16 | pass→pass | 14,486 | 7,044 | -51% | 1 | 1 | 0% | 2,463 | 1,605 | -35% | 0 | 0 | — |
case-17 | fail→pass | 19,319 | 4,598 | -76% | 1 | 1 | 0% | 3,012 | 1,261 | -58% | 0 | 0 | — |
case-18 | fail→pass | 21,828 | 5,904 | -73% | 1 | 1 | 0% | 3,154 | 1,683 | -47% | 0 | 0 | — |
case-19 | fail→pass | 16,725 | 8,214 | -51% | 1 | 1 | 0% | 2,212 | 1,714 | -23% | 0 | 0 | — |
case-20 | pass→pass | 17,931 | 2,767 | -85% | 1 | 1 | 0% | 2,468 | 1,043 | -58% | 0 | 0 | — |
case-21 | pass→pass | 17,291 | 17,048 | -1% | 1 | 1 | 0% | 2,454 | 2,772 | +13% | 0 | 0 | — |
case-22 | pass→fail | 28,536 | 8,547 | -70% | 1 | 1 | 0% | 5,569 | 1,067 | -81% | 0 | 0 | — |
case-23 | pass→fail | 19,999 | 7,306 | -63% | 1 | 1 | 0% | 2,899 | 912 | -69% | 0 | 0 | — |
case-24 | pass→fail | 26,937 | 17,425 | -35% | 1 | 1 | 0% | 3,891 | 1,170 | -70% | 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 18 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 0 percentage points is the difference between those two pass rates over the 18 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.