Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股融资融券/两融分析。当用户说"融资融券"、"两融"、"融资余额"、"融券"、"margin trading"、"杠杆资金"、"融资盘"、"XX的融资余额"、"两融风向"、"杠杆水平"时触发。基于 akshare 融资融券数据和 cn-stock-data 行情数据,分析全市场两融趋势、个股融资融券变化、杠杆资金方向判断。支持研报风格(formal)和快速解读风格(brief)。不适用于北向资金(用 a-share-northbound)或基金持仓(用 a-share-fund-holding)。
.claude/skills/aifinlab-a-share-margin-trading/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -57% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -9% | 0% |
bash# 融资融券数据(通过 akshare) # 全市场融资融券汇总 python -c "import akshare as ak; df=ak.stock_margin_sse(start_date='20260101'); print(df.to_json(orient='records', force_ascii=False))" # 个股融资融券明细 python -c "import akshare as ak; df=ak.stock_margin_detail_sse(date='20260315'); print(df.to_json(orient='records', force_ascii=False))" # 行情数据 SCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" python "$SCRIPTS/cn_stock_data.py" quote --code [CODE] python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [日期]
> akshare 接口名可能随版本变化,如接口不可用,通过 web 搜索获取融资融券数据。
获取融资余额/融券余额时间序列 + 个股行情数据。
| 维度 | formal | brief | |------|--------|-------| | 输出 | 完整两融分析报告 | 快速要点 | | 时间跨度 | 3-6个月趋势 | 近1-2周变化 | | 分析深度 | 市场+行业+个股三层 | 仅关注核心方向 | | 图表 | 融资余额走势图数据 | 无 |
详见 references/margin-trading-guide.md
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 | 34,213 | 21,490 | -37% | 1 | 1 | 0% | 5,235 | 1,639 | -69% | 0 | 0 | — |
case-02 | pass→fail | 20,080 | 9,564 | -52% | 1 | 1 | 0% | 3,365 | 1,322 | -61% | 0 | 0 | — |
case-08 | pass→pass | 12,654 | 18,131 | +43% | 1 | 1 | 0% | 1,805 | 3,422 | +90% | 0 | 0 | — |
case-09 | pass→pass | 11,793 | 8,473 | -28% | 1 | 1 | 0% | 1,921 | 2,127 | +11% | 0 | 0 | — |
case-10 | pass→pass | 17,326 | 7,209 | -58% | 1 | 1 | 0% | 2,597 | 1,839 | -29% | 0 | 0 | — |
case-15 | pass→pass | 13,322 | 9,720 | -27% | 1 | 1 | 0% | 1,759 | 2,098 | +19% | 0 | 0 | — |
case-03 | pass→pass | 7,567 | 2,716 | -64% | 1 | 1 | 0% | 1,183 | 1,257 | +6% | 0 | 0 | — |
case-04 | pass→pass | 9,880 | 4,073 | -59% | 1 | 1 | 0% | 1,447 | 1,463 | +1% | 0 | 0 | — |
case-05 | fail→pass | 16,035 | 2,824 | -82% | 1 | 1 | 0% | 2,845 | 1,168 | -59% | 0 | 0 | — |
case-06 | fail→pass | 7,464 | 4,277 | -43% | 1 | 1 | 0% | 1,170 | 1,650 | +41% | 0 | 0 | — |
case-07 | fail→pass | 17,305 | 2,951 | -83% | 1 | 1 | 0% | 2,765 | 1,195 | -57% | 0 | 0 | — |
case-11 | pass→pass | 13,818 | 16,501 | +19% | 1 | 1 | 0% | 2,003 | 3,141 | +57% | 0 | 0 | — |
case-12 | fail→pass | 14,467 | 13,541 | -6% | 1 | 1 | 0% | 2,157 | 2,942 | +36% | 0 | 0 | — |
case-13 | fail→pass | 18,447 | 11,446 | -38% | 1 | 1 | 0% | 2,783 | 2,526 | -9% | 0 | 0 | — |
case-14 | fail→fail | 9,315 | 9,715 | +4% | 1 | 1 | 0% | 1,493 | 2,097 | +40% | 0 | 0 | — |
case-16 | pass→pass | 7,386 | 5,060 | -31% | 1 | 1 | 0% | 1,101 | 1,626 | +48% | 0 | 0 | — |
case-17 | pass→pass | 10,446 | 5,309 | -49% | 1 | 1 | 0% | 1,572 | 1,611 | +2% | 0 | 0 | — |
case-18 | pass→pass | 19,191 | 18,126 | -6% | 1 | 1 | 0% | 2,780 | 3,650 | +31% | 0 | 0 | — |
case-19 | fail→pass | 14,052 | 4,162 | -70% | 1 | 1 | 0% | 2,199 | 1,285 | -42% | 0 | 0 | — |
case-20 | pass→fail | 21,768 | 11,510 | -47% | 1 | 1 | 0% | 3,112 | 1,438 | -54% | 0 | 0 | — |
case-21 | pass→pass | 9,318 | 9,428 | +1% | 1 | 1 | 0% | 1,678 | 2,345 | +40% | 0 | 0 | — |
case-22 | pass→pass | 7,767 | 7,321 | -6% | 1 | 1 | 0% | 1,578 | 2,121 | +34% | 0 | 0 | — |
case-23 | pass→pass | 13,118 | 14,129 | +8% | 1 | 1 | 0% | 1,921 | 2,781 | +45% | 0 | 0 | — |
case-24 | fail→pass | 9,520 | 1,860 | -80% | 1 | 1 | 0% | 1,288 | 1,053 | -18% | 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 21 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 +21 percentage points is the difference between those two pass rates over the 21 comparable cases. 2 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/28/2026 | 0% |
Other measured skills in the registry, with their headline benchmark lift.