Openclaw Security Audit

**Judgement** A strong, specialized security-audit skill for OpenClaw/Clawdbot deployments. It is well-scoped, read-only by default, and follows a clear, methodical checklist that is suitable for most Linux-based OpenClaw environments. **Key Risks / Limitations** - Relies on common Linux utilities (`ss`, `systemctl`, `journalctl`, `find`, etc.); coverage and reliability will degrade on stripped-down systems, non-Linux OSes, or containers without these tools. - Some checks (e.g., `find / -perm -4000`) can be relatively heavy on large systems, which may impact performance during the audit. - Focuses on configuration and host-level misconfigurations; it does not replace deeper application-specific security testing or network penetration testing. **Recommended Scenarios** - Use when you need a structured, repeatable hardening review of an OpenClaw/Clawdbot/Moltbot setup, especially prior to exposing it to the internet or to untrusted users. - Use for periodic re-audits after changing gateway config, adding new skills, or modifying exec/tool policies. - Particularly helpful for identifying unsafe public exposure of the gateway/control UI, weak exec policies, plaintext secrets in `~/.openclaw`, and risky or untrusted skills, along with concrete remediation steps.