stock-data-fetcher
获取A股数据(baostock)并缓存到本地CSV文件,避免MCP返回大量数据占用上下文。触发场景:(1)获取超过100条的K线数据 (2)需要多次查询同一股票数据 (3)需要用grep/awk分析数据 (4)用户提到"保存数据"或"缓存数据"
When & Why to Use This Skill
This Claude skill enables the retrieval of Chinese A-share market data via the Baostock library, featuring an automated local CSV caching mechanism. It is specifically designed to solve the 'context bloat' problem in LLMs by storing large datasets—such as extensive historical K-lines and multi-year financial reports—locally, allowing for efficient data manipulation and analysis without exhausting the model's token limit.
Use Cases
- Case 1: Analyzing long-term historical trends by fetching over 100 periods of K-line data for technical analysis while keeping the conversation context lean.
- Case 2: Conducting deep fundamental research by caching and querying quarterly financial indicators like ROE, net profit margins, and growth rates.
- Case 3: Improving efficiency for repeated stock queries by utilizing local CSV caches to avoid redundant API calls and reduce latency.
- Case 4: Performing high-performance data filtering and signal detection using local shell tools like grep and awk on cached market datasets.
| name | stock-data-fetcher |
|---|---|
| description | 获取A股数据(baostock)并缓存到本地CSV文件,避免MCP返回大量数据占用上下文。触发场景:(1)获取超过100条的K线数据 (2)需要多次查询同一股票数据 (3)需要用grep/awk分析数据 (4)用户提到"保存数据"或"缓存数据" |
Stock Data Fetcher
获取A股数据并保存到本地CSV,通过grep/read按需查看,避免上下文膨胀。
快速开始
# 获取K线数据(默认1年,前复权,保存到 ./data/cache)
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138
# 获取近2年数据
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138 --days 730
# 指定日期范围
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.600000 --start 2024-01-01 --end 2024-12-31
# 获取财务数据
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138 --type profit --year 2024 --quarter 3
# 指定缓存目录
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138 --cache-dir ./my_data
# 或通过环境变量设置缓存目录
STOCK_CACHE_DIR=./my_data uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138
缓存目录
优先级: --cache-dir > STOCK_CACHE_DIR 环境变量 > ./data/cache
{cache_dir}/
sh_601138/
daily_2024-01-01_2024-12-31.csv # K线
profit_2024Q3.csv # 盈利
growth_2024Q3.csv # 成长
数据类型
| 类型 | 参数 | 说明 |
|---|---|---|
| kline | --freq d/w/m |
K线,支持日/周/月/分钟 |
| profit | --year --quarter |
盈利能力(ROE/净利率等) |
| growth | --year --quarter |
成长能力(YOY增长率) |
| balance | --year --quarter |
偿债能力(资产负债率等) |
| cashflow | --year --quarter |
现金流 |
| dupont | --year --quarter |
杜邦分析 |
| dividend | --year |
分红数据 |
查看缓存数据
# 列出所有缓存
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py --list
# 列出指定股票缓存
uv run python .claude/skills/stock-data-fetcher/scripts/fetch_stock_data.py sh.601138 --list
# 查看最近10条
tail -10 data/cache/sh_601138/daily_*.csv
# 搜索特定月份
grep "2024-09" data/cache/sh_601138/daily_*.csv
# 筛选涨幅>5%的日期
awk -F',' 'NR==1 || $13>5' data/cache/sh_601138/daily_*.csv
K线字段
| 字段 | 说明 |
|---|---|
| date | 日期 |
| open/high/low/close | 开高低收 |
| volume | 成交量(股) |
| amount | 成交额(元) |
| turn | 换手率(%) |
| pctChg | 涨跌幅(%) |
| peTTM | 滚动市盈率 |
| pbMRQ | 市净率 |
依赖
首次使用需安装: uv add baostock pandas