VEXT Monitor
Real-time behavioral monitoring for your OpenClaw installation. Watches for unauthorized file access, suspicious network activity, identity file tampering, and rogue processes.
Usage
- "Monitor my skills for suspicious activity"
- "Check if any skills have modified my SOUL.md"
- "Run a security monitor check"
- "Watch for unauthorized changes"
How It Works
- Hashes SOUL.md, MEMORY.md, AGENTS.md, and openclaw.json at startup
- Monitors for file modifications to identity and config files
- Checks for sensitive file access (.env, SSH keys, cookies, keychains)
- Scans for suspicious network connections
- Detects new background processes that may have been spawned by skills
- Logs all findings to ~/.openclaw/vext-shield/monitor.log
Rules
- Only perform read-only monitoring — never modify monitored files
- Log all findings with timestamps
- Do not interfere with legitimate skill operations
- Do not transmit monitoring data externally
Safety
- This skill only reads files and system state — no modifications
- No network requests are made
- Monitoring data stays local in ~/.openclaw/vext-shield/
- Can be run as a one-shot check or on a schedule