Install any skill in seconds. Free to start, no credit card required.
Get Started Free →AKShare财经数据接口库封装,提供股票、期货、期权、基金、外汇、债券、指数、加密货币等金融产品的基本面数据、实时和历史行情数据、衍生数据。
.claude/skills/aifinlab-akshare-finance/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -48% | 0% |
bash# 安装依赖 pip install akshare pandas # 测试安装 python -c "import akshare; print(akshare.__version__)"
pythonimport akshare as ak # A股实时行情 stock_zh_a_spot_em() # 东方财富A股 # 股票K线数据 stock_zh_kline(symbol="000001", period="daily", adjust="qfq") # 港股行情 stock_hk_spot_em() # 港股实时 # 美股 stock_us_spot() # 美股实时
python# GDP数据 macro_china_gdp() # 中国GDP # CPI通胀 macro_china_cpi() # 中国CPI # PMI采购经理指数 macro_china_pmi() # 中国PMI # 货币供应量 macro_china_m2() # M2广义货币
python# 币种列表 crypto_binance_symbols() # 币安交易对 # 实时价格 crypto_binance_btc_usdt_spot() # BTC/USDT # K线数据 crypto_binance_btc_usdt_kline(period="daily")
python# 外汇汇率 forex_usd_cny() # 美元兑人民币 # 贵金属 metals_shibor() # 上海银行间拆借利率 # 金银价格 metals_gold() # 国际金价
python# 股票基本面 stock_fundamental(symbol="000001") # 基本面数据 # 估值指标 stock_valuation(symbol="000001") # PE、PB等 # 盈利能力 stock_profit_em(symbol="000001")
pythonimport akshare as ak import pandas as pd # 监控自选股 tickers = ["000001", "000002", "600519"] for ticker in tickers: df = ak.stock_zh_kline(symbol=ticker, period="daily", adjust="qfq", start_date="20240101") latest = df.iloc[-1] print(f"{ticker}: 收盘价={latest['close']}, 涨跌幅={latest['pct_chg']}%")
python# A股大盘 index_zh_a_spot() # 大盘指数 # 涨跌幅排行 stock_zh_a_spot_em()[['代码', '名称', '涨跌幅']].sort_values('涨跌幅', ascending=False)
默认返回Pandas DataFrame,可直接处理:
pythondf = ak.stock_zh_a_spot_em() print(df.head()) # 查看前5行 print(df.columns) # 查看列名 df.to_csv("data.csv") # 保存CSV
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 | 12,856 | 12,526 | -3% | 1 | 1 | 0% | 2,683 | 2,953 | +10% | 0 | 0 | — |
case-02 | fail→fail | 18,485 | 7,947 | -57% | 1 | 1 | 0% | 3,784 | 2,642 | -30% | 0 | 0 | — |
case-20 | pass→pass | 12,421 | 11,269 | -9% | 1 | 1 | 0% | 2,135 | 2,663 | +25% | 0 | 0 | — |
case-21 | pass→fail | 18,596 | 19,818 | +7% | 1 | 1 | 0% | 3,342 | 4,520 | +35% | 0 | 0 | — |
case-03 | pass→pass | 13,822 | 7,318 | -47% | 1 | 1 | 0% | 2,096 | 2,030 | -3% | 0 | 0 | — |
case-04 | pass→pass | 8,962 | 9,744 | +9% | 1 | 1 | 0% | 1,585 | 2,761 | +74% | 0 | 0 | — |
case-05 | fail→fail | 7,685 | 7,101 | -8% | 1 | 1 | 0% | 1,073 | 2,173 | +103% | 0 | 0 | — |
case-06 | pass→pass | 11,759 | 5,554 | -53% | 1 | 1 | 0% | 2,087 | 1,946 | -7% | 0 | 0 | — |
case-07 | fail→fail | 9,455 | 6,436 | -32% | 1 | 1 | 0% | 1,573 | 2,044 | +30% | 0 | 0 | — |
case-08 | fail→pass | 9,336 | 3,102 | -67% | 1 | 1 | 0% | 1,664 | 1,373 | -17% | 0 | 0 | — |
case-09 | fail→pass | 10,147 | 7,094 | -30% | 1 | 1 | 0% | 1,770 | 2,229 | +26% | 0 | 0 | — |
case-10 | fail→pass | 9,354 | 6,826 | -27% | 1 | 1 | 0% | 1,364 | 1,858 | +36% | 0 | 0 | — |
case-11 | fail→pass | 10,553 | 3,285 | -69% | 1 | 1 | 0% | 1,802 | 1,493 | -17% | 0 | 0 | — |
case-12 | fail→fail | 7,613 | 8,714 | +14% | 1 | 1 | 0% | 1,285 | 2,667 | +108% | 0 | 0 | — |
case-13 | fail→pass | 18,270 | 2,192 | -88% | 1 | 1 | 0% | 2,478 | 1,287 | -48% | 0 | 0 | — |
case-14 | fail→fail | 7,796 | 10,032 | +29% | 1 | 1 | 0% | 1,294 | 2,839 | +119% | 0 | 0 | — |
case-15 | pass→pass | 6,260 | 8,505 | +36% | 1 | 1 | 0% | 1,087 | 2,565 | +136% | 0 | 0 | — |
case-16 | fail→fail | 8,305 | 10,464 | +26% | 1 | 1 | 0% | 1,190 | 2,418 | +103% | 0 | 0 | — |
case-17 | fail→fail | 9,987 | 9,762 | -2% | 1 | 1 | 0% | 1,897 | 2,278 | +20% | 0 | 0 | — |
case-18 | fail→fail | 10,532 | 8,509 | -19% | 1 | 1 | 0% | 2,051 | 2,392 | +17% | 0 | 0 | — |
case-19 | pass→pass | 11,583 | 11,174 | -4% | 1 | 1 | 0% | 1,798 | 2,548 | +42% | 0 | 0 | — |
case-22 | pass→pass | 12,142 | 10,524 | -13% | 1 | 1 | 0% | 2,271 | 3,106 | +37% | 0 | 0 | — |
case-23 | pass→pass | 8,741 | 3,995 | -54% | 1 | 1 | 0% | 1,283 | 1,617 | +26% | 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. The headline lift of +17 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.