Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →资金流向Skill - 追踪主力资金、散户资金、北向资金、板块资金流向 via AkShare/东方财富
.claude/skills/aifinlab-akshare-capital/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 62% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -20% | 0% |
| 属性 | 内容 | |:---|:---| | 名称 | akshare-capital | | 版本 | 1.1.0 | | 分类 | 资金流向 | | 状态 | ✅ 已上线 | | 维护者 | FinClaw Core Team | | 最后更新 | 2026-03-19 |
资金流向Skill，提供个股资金流向、板块资金流向、主力资金排行、北向资金（陆股通）明细等数据。追踪超大单/大单/中单/小单资金动向，识别主力建仓/出货信号。
| 用户输入 | 识别意图 | 调用函数 | |:---|:---|:---| | 茅台资金流向如何？ | capital_flow | capital_flow.py 600519 | | 主力资金流入哪些板块？ | capital_board | capital_board.py | | 北向资金今天买了什么？ | capital_north | capital_north.py | | 主力资金排名 | capital_main | capital_main.py |
| 数据类型 | 主要来源 | 备用来源 | 认证要求 | |:---|:---|:---|:---:| | 个股资金流向 | AkShare-东方财富 | - | 无需 | | 板块资金流向 | AkShare-东方财富 | - | 无需 | | 北向资金 | AkShare-交易所 | - | 无需 | | 主力资金排行 | AkShare-东方财富 | - | 无需 |
| 类型 | 定义 | 代表意义 | |:---|:---|:---| | 超大单 | >100万股或>100万元 | 机构主力、大资金建仓/出货 | | 大单 | 20-100万股或20-100万元 | 大资金、基金调仓 | | 中单 | 4-20万股或4-20万元 | 中等资金、私募/大户 | | 小单 | <4万股或<4万元 | 散户资金 |
bash python scripts/capital_flow.py 600519 python scripts/capital_flow.py 300750
bash python scripts/capital_board.py python scripts/capital_board.py --limit 20
bash python scripts/capital_north.py
bash python scripts/capital_main.py python scripts/capital_main.py --market sh --limit 30
| 脚本名 | 功能 | 入口点 | |:---|:---|:---:| | capital_flow.py | 个股资金流向 | ✅ | | capital_board.py | 板块资金流向 | ✅ | | capital_north.py | 北向资金明细 | ✅ | | capital_main.py | 主力资金排行 | ✅ |
| 情形 | 超大单 | 大单 | 中单 | 小单 | 解读 | |:---|:---:|:---:|:---:|:---:|:---| | 主力建仓 | +++ | ++ | - | --- | 机构买入，散户卖出 | | 主力出货 | --- | -- | + | +++ | 机构卖出，散户接盘 | | 一致看多 | +++ | ++ | + | ++ | 各资金集体流入 | | 一致看空 | --- | -- | - | -- | 各资金集体流出 | | 震荡洗盘 | +- | -+ | +- | +- | 资金分歧，震荡为主 |
| 北向资金 | 信号 | 意义 | |:---|:---:|:---| | 连续3日净流入 >50亿 | 🟢 强烈看多 | 外资大幅加仓 | | 连续3日净流出 >50亿 | 🔴 强烈看空 | 外资大幅减仓 | | 单日净流入 >100亿 | 🟢 积极信号 | 外资抢筹 | | 单日净流出 >100亿 | 🔴 警示信号 | 外资逃离 | | 小幅波动 ±20亿 | ⚪ 中性 | 正常波动 |
本Skill所有输出数据将按以下格式标注来源：
markdown--- 📊 **数据来源**: AkShare-东方财富 ⏱️ **数据时间**: 2026-03-19 10:30:15 📌 **统计口径**: 超大单(≥100万)/大单(20-100万)/中单(4-20万)/小单(<4万) 🔗 **原始来源**: 东方财富资金流向 🔧 **分析工具**: FinClaw v1.0
akshare>=1.10.0
pandas>=1.3.0
pyyaml>=5.4.0| 版本 | 日期 | 变更内容 | |:---|:---:|:---| | 1.1.0 | 2026-03-19 | 符合FinClaw数据规范v1.0 | | 1.0.0 | 2026-03-13 | 初始版本 |
finclaw/config/data_source_config.yaml本Skill遵循 FinClaw 数据规范 v1.0 | 数据来源强制标注 | 禁止训练数据编造
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 14,155 | 11,444 | -19% | 1 | 1 | 0% | 2,404 | 2,338 | -3% | 0 | 0 | — |
case-02 | fail→fail | 17,174 | 6,095 | -65% | 1 | 1 | 0% | 2,512 | 2,147 | -15% | 0 | 0 | — |
case-03 | fail→fail | 21,339 | 23,610 | +11% | 1 | 1 | 0% | 3,050 | 5,002 | +64% | 0 | 0 | — |
case-04 | pass→pass | 15,582 | 23,203 | +49% | 1 | 1 | 0% | 2,951 | 6,111 | +107% | 0 | 0 | — |
case-05 | pass→pass | 17,956 | 27,370 | +52% | 1 | 1 | 0% | 3,340 | 6,090 | +82% | 0 | 0 | — |
case-06 | pass→fail | 11,203 | 18,882 | +69% | 1 | 1 | 0% | 1,718 | 4,458 | +159% | 0 | 0 | — |
case-07 | pass→pass | 20,298 | 20,778 | +2% | 1 | 1 | 0% | 2,613 | 4,473 | +71% | 0 | 0 | — |
case-08 | pass→pass | 23,275 | 23,961 | +3% | 1 | 1 | 0% | 2,446 | 4,250 | +74% | 0 | 0 | — |
case-09 | pass→pass | 21,161 | 12,459 | -41% | 1 | 1 | 0% | 2,908 | 3,944 | +36% | 0 | 0 | — |
case-10 | fail→pass | 19,151 | 17,479 | -9% | 1 | 1 | 0% | 2,918 | 4,719 | +62% | 0 | 0 | — |
case-11 | fail→pass | 13,332 | 3,955 | -70% | 1 | 1 | 0% | 2,221 | 2,548 | +15% | 0 | 0 | — |
case-12 | pass→pass | 13,548 | 6,175 | -54% | 1 | 1 | 0% | 2,263 | 2,694 | +19% | 0 | 0 | — |
case-13 | pass→pass | 15,878 | 3,556 | -78% | 1 | 1 | 0% | 1,891 | 2,459 | +30% | 0 | 0 | — |
case-14 | pass→pass | 21,444 | 12,079 | -44% | 1 | 1 | 0% | 2,564 | 3,407 | +33% | 0 | 0 | — |
case-15 | pass→pass | 20,033 | 4,368 | -78% | 1 | 1 | 0% | 2,510 | 2,472 | -2% | 0 | 0 | — |
case-16 | pass→pass | 13,482 | 3,846 | -71% | 1 | 1 | 0% | 2,004 | 2,437 | +22% | 0 | 0 | — |
case-17 | pass→pass | 20,460 | 22,154 | +8% | 1 | 1 | 0% | 2,910 | 4,825 | +66% | 0 | 0 | — |
case-18 | fail→pass | 12,974 | 3,087 | -76% | 1 | 1 | 0% | 2,777 | 2,286 | -18% | 0 | 0 | — |
case-19 | fail→pass | 18,544 | 5,041 | -73% | 1 | 1 | 0% | 3,464 | 2,541 | -27% | 0 | 0 | — |
case-20 | fail→pass | 15,980 | 2,521 | -84% | 1 | 1 | 0% | 2,816 | 2,264 | -20% | 0 | 0 | — |
case-21 | fail→pass | 35,051 | 7,635 | -78% | 1 | 1 | 0% | 3,155 | 3,254 | +3% | 0 | 0 | — |
case-22 | fail→pass | 15,265 | 6,967 | -54% | 1 | 1 | 0% | 2,245 | 3,079 | +37% | 0 | 0 | — |
case-23 | pass→pass | 21,618 | 18,473 | -15% | 1 | 1 | 0% | 2,598 | 4,048 | +56% | 0 | 0 | — |
case-24 | pass→pass | 17,584 | 18,534 | +5% | 1 | 1 | 0% | 2,193 | 4,364 | +99% | 0 | 0 | — |
case-25 | pass→pass | 16,964 | 21,226 | +25% | 1 | 1 | 0% | 2,399 | 4,772 | +99% | 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. 25 cases were attempted, and 23 counted toward the lift figure. The other 2 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 +24 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is 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.