A股市場資料

Use this skill to query a fixed whitelist of AKShare market datasets through the bundled CLI.

## Workflow

1. Convert relative dates to absolute `YYYYMMDD` before calling the script.
2. Map the user's request to one of the supported subcommands in [`references/datasets.md`](references/datasets.md).
3. Run only the bundled script with OpenClaw `exec`:

```bash
python "{baseDir}/scripts/query_akshare.py" <subcommand> [args...]
```

4. Read the JSON envelope from stdout.
5. Reply in Chinese with a short summary plus the most relevant values from the returned `rows`.

## Hard Rules

- Use only `python "{baseDir}/scripts/query_akshare.py" ...`.
- Do not paste the user's raw request into shell pipelines.
- Do not generate ad-hoc Python snippets to call AKShare directly.
- Do not guess when a symbol match is ambiguous; surface the returned candidates.
- Keep dates in `YYYYMMDD`.
- Treat missing `akshare` as an environment issue and relay the install command from the error payload.

## Supported Commands

- `market-overview`
  - Return A-share market breadth, major indices, top 10 gainers, and top 10 losers.
- `stock-quote <symbol>`
  - Real-time A-share quote by code or exact name.
- `stock-history <symbol> [--start-date YYYYMMDD] [--end-date YYYYMMDD] [--period daily|weekly|monthly] [--adjust none|qfq|hfq] [--limit N]`
- `stock-profile <symbol>`
  - Company and security profile from Eastmoney.
- `index-quote <symbol>`
  - Real-time China index quote by code or exact name.
- `index-history <symbol> [--start-date YYYYMMDD] [--end-date YYYYMMDD] [--period daily|weekly|monthly] [--limit N]`
- `fund-quote <symbol>`
  - Open-end mutual fund daily NAV snapshot by code or exact name.
- `fund-history <symbol> [--indicator unit_nav|acc_nav|acc_return] [--period 1m|3m|6m|1y|3y|5y|ytd|all] [--limit N]`
- `macro-series <alias>`
  - Supported aliases: `china_cpi`, `china_pmi`, `china_rmb`.
- `macro-calendar [--date YYYYMMDD] [--limit N]`
- `news-flash [--scope all|important] [--limit N]`

## Command Examples

```bash
python "{baseDir}/scripts/query_akshare.py" market-overview
python "{baseDir}/scripts/query_akshare.py" stock-quote 000001
python "{baseDir}/scripts/query_akshare.py" stock-history 600519 --start-date 20260101 --end-date 20260313
python "{baseDir}/scripts/query_akshare.py" index-history 000300 --limit 60
python "{baseDir}/scripts/query_akshare.py" fund-history 000001 --indicator unit_nav --limit 60
python "{baseDir}/scripts/query_akshare.py" macro-series china_pmi
python "{baseDir}/scripts/query_akshare.py" macro-calendar --date 20260313
python "{baseDir}/scripts/query_akshare.py" news-flash --limit 10
```

## Output Contract

Every success response is a JSON object with:

- `ok`
- `dataset`
- `akshare_function`
- `resolved`
- `params`
- `columns`
- `rows`
- `row_count`
- `truncated`
- `as_of`

Every failure response is a JSON object with:

- `ok: false`
- `error.type`
- `error.message`
- optional `error.details`

Error types are limited to:

- `missing_dependency`
- `invalid_argument`
- `unsupported_query`
- `ambiguous_symbol`
- `runtime_error`

## Summary Style

- Lead with the direct answer in Chinese.
- Mention the symbol or dataset that was resolved.
- For quotes, include latest price and change fields when present.
- For history, describe the date range covered by the returned rows.
- For market overview, mention breadth and the main index moves.
- For macro/news results, surface the latest or most important rows instead of dumping the whole payload.