Menu
灵台量化数据管道——股票/加密币行情采集、清洗、存储与质量检查。含 Yahoo Finance + Kraken (CCXT) 双数据源，支持日线/1h 自动增量更新。
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Configure the Feishu (Lark) bot addon for this agent — read this when the human asks to set up a Feishu bot.
Configure the IMAP email addon for this agent — read this when the human asks to set up email.
Configure the Telegram bot addon for this agent — read this when the human asks to set up a Telegram bot.
Configure the WeChat addon for this agent — read this when the human asks to set up WeChat.
Reference manual for the `avatar` tool — debugging, advanced patterns, best practices, and worked examples. Read when you need depth beyond the tool description.
Reference manual for the `bash` tool — debugging, advanced patterns, best practices, and worked examples. Read when you need depth beyond the tool description.
| name | quant-data |
| description | 灵台量化数据管道——股票/加密币行情采集、清洗、存储与质量检查。含 Yahoo Finance + Kraken (CCXT) 双数据源，支持日线/1h 自动增量更新。 |
| version | 1.0.0 |
| agent | data_engineer |
| created | "2026-04-25T00:00:00.000Z" |
股票（Yahoo Finance）与加密货币（Kraken/CCXT）行情数据的采集、清洗、存储工具集。
| 源 | 类型 | 标的 | 限制 |
|---|---|---|---|
| Yahoo Finance | 美股日线/1h | SPY, QQQ, IWM, GLD, TLT, AAPL, MSFT, GOOGL, AMZN, NVDA | 日内深度 ~90 天 |
| Kraken (CCXT) | 加密币日线/1h | BTC/USD, ETH/USD | 1h 回溯上限 ~720 条 |
| Binance | ❌ 不可用 | — | 美国地区 HTTP 451 |
data/raw/{TICKER}/ ← Parquet 文件，按标的分目录
data/processed/ ← 清洗后数据（待建）
data/features/ ← 特征工程输出（待建）
文件命名：{TICKER}_{freq}_{开始日期}_{结束日期}.parquet
cd /Users/huangzesen/work/lingtai-projects/quant_company
python3 scripts/ingest_all.py --status
# 全量日线 + 日内
python3 scripts/ingest_all.py --all
# 仅日线增量更新
python3 scripts/ingest_all.py --daily
# 仅日内 1h
python3 scripts/ingest_all.py --intraday
# 指定标的
python3 scripts/ingest_all.py --daily --tickers SPY QQQ AAPL
python3 scripts/ingest_all.py --quality
python3 scripts/ingest_all.py --quality daily
python3 scripts/ingest_all.py --quality intraday
shared_lib/data_fetcher.py — 数据获取层from shared_lib.data_fetcher import fetch_yahoo, fetch_ccxt, load_local, list_local_data
# 采集 Yahoo 数据
df = fetch_yahoo('SPY', interval='1d', start=..., end=...)
# 采集 Kraken 数据
df = fetch_ccxt('BTC/USD', exchange='kraken', timeframe='1d', limit=500)
# 加载本地数据（自动合并去重排序）
df = load_local('SPY')
# 列出本地文件
files = list_local_data('SPY')
shared_lib/features.py — 技术指标from shared_lib.features import add_all_indicators
df = add_all_indicators(df) # 返回含 SMA/MACD/RSI/布林带/ATR/OBV 的 DataFrame
ingest_all.py --daily 会：
设置每日收盘后运行：
# 每天 16:30 ET 运行（美股收盘后半小时）
30 20 * * 1-5 cd /path/to/quant_company && python3 scripts/ingest_all.py --daily
/opt/anaconda3/bin/python3