Your OpenClaw setup runs 24/7. But what watches the watchers?
Auto-Watchdog monitors everything and fixes what it can — silently. You only hear about real problems.
Checks all cron jobs every heartbeat:
consecutiveErrors > 0 → immediate alertMonitors critical processes by log freshness (not just PID):
openclaw gateway status every heartbeat## 🔍 Health Check (silent = good)
### Crons
- `cron list` → check consecutiveErrors > 0 → alert
- Frequent crons not running >2 hours → alert
### Processes
- Check [your process] log freshness < [X] minutes
- If stale → restart and alert
### Gateway
- `openclaw gateway status` → alert if down
### Disk
- Check log sizes > 10MB → rotate
- Check workspace size > 1GB → alert
Create a VBS wrapper for zero-flash execution:
' guardian-silent.vbs — zero flash process monitor
Set oShell = CreateObject("WScript.Shell")
oShell.Run "powershell.exe -NonInteractive -WindowStyle Hidden -ExecutionPolicy Bypass -File ""C:\path\to\guardian.ps1""", 0, True
Register as Task Scheduler job running every 1-5 minutes.
# /etc/systemd/system/openclaw-watchdog.service
[Service]
ExecStart=/usr/bin/node /path/to/guardian.js
Restart=always
RestartSec=60
> Monitor by output freshness, not PID.
> A process can be "alive" but stuck. Check its log timestamp.
> Fix first, alert second.
> If you can restart it automatically, do it. Only alert for things that need human intervention.
> Silent = healthy.
> No news is good news. Only speak up when something breaks.
Built for a 24/7 autonomous trading system with:
共 1 个版本