About this skill

Problem to Solve

Manual price watching is repetitive and easy to miss thresholds or percentage moves. This skill targets configurable A-share stock alerts, such as notifying when a price enters a range or daily change crosses preset percentages.

How It Works

It reads market data from public endpoints via akshare, with no API Key required. The main workflow includes:

Add monitoring: supports threshold rules and quick interactive setup, e.g. quick_add("000001", "12~15") and quick_add("600519", "+3%,-2%").
Review and manage: use list_watchlist() to inspect the current watchlist and adjust entries.
Trigger checks: use run_check() to fetch real-time prices, evaluate alert conditions, and review alert history.

Configuration and history are stored locally in files such as ~/.hermes/stock_watchlist.json and ~/.hermes/stock_alerts.json, making rules reusable and auditable.

Boundaries

Data availability depends on public market APIs and network conditions, and non-trading hours may return no data. Stock codes should be 6-digit numeric codes, and dependencies such as akshare and pandas must be present. It is better suited to threshold monitoring and alert checks rather than guaranteed real-time broker push or trading execution.

Screenshots

Use Cases

Set a 12-15 range alert for 000001, then review rules with `list_watchlist()`.
Add +3% and -2% alerts for 600519 and check triggers during trading hours.
Run `run_check()` to evaluate configured stocks against thresholds and record results.
Review `~/.hermes/stock_alerts.json` to trace a stock's past trigger records.

Best For

A-share individual investors who want automated reminders for price ranges like 12-15.
Quant analysts who need stored trigger results and historical alert review.
Data engineers who schedule agent checks every 5 minutes during trading hours.
Research assistants managing multiple watchlists for codes like 000001 and 600519.