OpenClaw Healthcheck

Run a lightweight operational and security review for an OpenClaw deployment.

What this skill checks

• gateway reachability and process status
• exposed listeners / suspicious open ports
• risky config patterns in the running OpenClaw config
• browser session / relay surface hints
• recent log errors worth operator attention
• update / runtime hygiene signals

When to use

Use this skill when a user asks:

• "run an OpenClaw health check"
• "audit my OpenClaw security"
• "is this OpenClaw deployment safe"
• "check my OpenClaw server exposure"
• "review my OpenClaw setup before production"
• "why does my OpenClaw runtime feel unhealthy"

Workflow

1. Confirm target machine / workspace.
2. Run the bundled healthcheck script.
3. Review score, findings, and recommendations.
4. If findings are high-risk, stop and fix before wider rollout.

Command

node {baseDir}/scripts/healthcheck.mjs

Output format

The script returns JSON with:

• score
• verdict
• findings
• recommendations
• evidence

Verdicts

• PASS — no major operational or security issue detected in this lightweight pass
• WARN — usable but risky or incomplete
• FAIL — high-risk exposure or broken runtime detected

Important limits

• This is a lightweight operator check, not a full penetration test.
• A PASS result does not guarantee safety.
• Human review is still required for internet-exposed systems.

References

• {baseDir}/references/checklist.md