Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股行业资金流向/板块主力资金分析。当用户说"行业资金流"、"sector fund flow"、"板块资金"、"哪个行业有资金流入"、"资金流去哪个行业"时触发。量化分析各行业资金流向。支持formal和brief风格。
.claude/skills/aifinlab-a-share-sector-fund-flow/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -57% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -44% | 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]
按行业市值归一化资金流入（每万亿市值净流入）
| 维度 | formal | brief | |------|--------|-------| | 行业排名 | 全行业资金流表 | Top/Bottom 5 | | 动量分析 | 资金流动量排名 | 加速流入板块 | 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-06 | pass→fail | 19,251 | 55,081 | +186% | 1 | 1 | 0% | 2,838 | 1,029 | -64% | 0 | 0 | — |
case-13 | fail→pass | 15,298 | 3,688 | -76% | 1 | 1 | 0% | 2,089 | 908 | -57% | 0 | 0 | — |
case-19 | fail→pass | 14,278 | 6,225 | -56% | 1 | 1 | 0% | 2,224 | 1,436 | -35% | 0 | 0 | — |
case-18 | fail→pass | 10,139 | 5,301 | -48% | 1 | 1 | 0% | 1,620 | 1,225 | -24% | 0 | 0 | — |
case-01 | fail→fail | 19,564 | 11,361 | -42% | 1 | 1 | 0% | 3,015 | 913 | -70% | 0 | 0 | — |
case-02 | fail→fail | 27,403 | 27,653 | +1% | 1 | 1 | 0% | 3,559 | 5,004 | +41% | 0 | 0 | — |
case-03 | fail→pass | 22,865 | 33,710 | +47% | 1 | 1 | 0% | 2,957 | 3,822 | +29% | 0 | 0 | — |
case-04 | pass→fail | 28,383 | 7,641 | -73% | 1 | 1 | 0% | 5,609 | 842 | -85% | 0 | 0 | — |
case-05 | fail→fail | 14,771 | 15,343 | +4% | 1 | 1 | 0% | 1,939 | 2,755 | +42% | 0 | 0 | — |
case-07 | pass→pass | 16,817 | 7,172 | -57% | 1 | 1 | 0% | 2,079 | 1,351 | -35% | 0 | 0 | — |
case-08 | fail→pass | 18,222 | 6,242 | -66% | 1 | 1 | 0% | 2,326 | 1,295 | -44% | 0 | 0 | — |
case-09 | pass→pass | 16,027 | 3,259 | -80% | 1 | 1 | 0% | 2,543 | 939 | -63% | 0 | 0 | — |
case-10 | fail→pass | 35,475 | 16,464 | -54% | 1 | 1 | 0% | 3,321 | 2,531 | -24% | 0 | 0 | — |
case-11 | pass→pass | 34,489 | 19,504 | -43% | 1 | 1 | 0% | 3,358 | 3,234 | -4% | 0 | 0 | — |
case-12 | pass→pass | 9,253 | 3,136 | -66% | 1 | 1 | 0% | 1,376 | 801 | -42% | 0 | 0 | — |
case-14 | fail→pass | 13,320 | 5,112 | -62% | 1 | 1 | 0% | 2,332 | 1,249 | -46% | 0 | 0 | — |
case-15 | pass→pass | 22,277 | 18,967 | -15% | 1 | 1 | 0% | 2,724 | 3,118 | +14% | 0 | 0 | — |
case-16 | pass→pass | 15,019 | 3,298 | -78% | 1 | 1 | 0% | 2,017 | 956 | -53% | 0 | 0 | — |
case-17 | fail→pass | 20,535 | 37,268 | +81% | 1 | 1 | 0% | 2,909 | 2,995 | +3% | 0 | 0 | — |
case-20 | pass→pass | 20,404 | 19,743 | -3% | 1 | 1 | 0% | 2,623 | 3,235 | +23% | 0 | 0 | — |
case-21 | fail→pass | 19,034 | 19,519 | +3% | 1 | 1 | 0% | 2,914 | 3,114 | +7% | 0 | 0 | — |
case-22 | pass→pass | 18,964 | 19,203 | +1% | 1 | 1 | 0% | 2,502 | 2,748 | +10% | 0 | 0 | — |
case-23 | pass→pass | 15,726 | 14,677 | -7% | 1 | 1 | 0% | 2,321 | 2,462 | +6% | 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. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 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 +30 percentage points is the difference between those two pass rates over the 20 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.