Tradebot Ops

Monitor trading bot health by detecting stalls, verifying LIVE status and data freshness, restarting safely if needed, and summarizing status clearly.

Purpose

Operate and monitor the trading bot safely: detect stalls, verify LIVE/halting state, validate data freshness, restart cleanly, and produce human-readable health summaries.

Use when

• Bot shows LIVE but no updates (stale heartbeat/bars)
• Chart frozen / signals file not updating
• Need to answer: “Is it running? Is it safe? Why no trades?”

Inputs

• dist/out/live_heartbeat_*.json
• dist/out/live_signals_*.csv
• dist/out/live_trades_*.csv (if present)
• UI endpoints: /api/trading/*

Outputs

• One-paragraph health summary
• If unhealthy: one action (restart/clear stale) + verification
• Audit log entry

Safety rails

• Never raise risk_pct automatically.
• Prefer restart/self-heal over loosening risk.

Checklist

1. Confirm UI server reachable /api/status.
2. Read heartbeat: mode/status/halted/in_position/last_bar_ts.
3. Freshness: _hb_age_s, _bar_age_s.
4. Confirm signals file mtime is recent.
5. If stale: stop bot PID (runtime or heartbeat pid) → apply+restart.
6. Re-check last_bar_ts advances.