fetch-stock-daily

Fetch Stock Daily JSON

Use this skill for the publishable Node.js path stock input -> symbol resolution -> daily history fetch -> raw JSON archive.

Quick Start

Run the Node.js helper from the repo root:

node skills/fetch-stock-daily-json/scripts/fetch_daily_json.js 贵州茅台
node skills/fetch-stock-daily-json/scripts/fetch_daily_json.js 600519 --years 3
node skills/fetch-stock-daily-json/scripts/fetch_daily_json.js 宁德时代 --start-date 20240101 --end-date 20260329 --adjust qfq

The script prints a JSON summary with the resolved symbol, row count, requested window, actual trade-date bounds, and saved file path.

Workflow

1. Accept repo-compatible inputs only: 6-digit code, 600519.SH or 300750.SZ, or a Chinese stock name. Do not use sh600519.
2. Prefer the bundled Node.js script instead of re-implementing the workflow in the task.
3. Resolve stock names through the Eastmoney A-share list API and cache the code-name map locally.
4. Fetch daily bars from the Eastmoney historical kline API with daily period and qfq/hfq/none adjustment.
5. Save the archived rows under data/raw/eastmoney/daily_history/YYYYMMDD/_.json.
6. Default to years=5 and adjust=qfq when the user does not provide a date window.
7. Stop after JSON archive generation unless the user explicitly asks for downstream parsing, analysis, or DB sync.

Return Format

Return the script's JSON summary and call out:

• resolved_symbol
• rows
• trade_date_min
• trade_date_max
• output_path

Failure Handling

• If name resolution fails, ask for a valid 6-digit A-share code or Chinese stock name.
• If the fetch returns zero rows, keep the archive file and report rows: 0.
• If you need the JS-side API details or the parity notes versus the Python workflow, read references/lobsterclaw-workflow.md.