AI Agent Hub
Back to skills
OpenClaw Host Security Check and Hardening icon

OpenClaw Host Security Check and Hardening

IT Ops & Security Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Follow https://skillhub.cn/install/skillhub.md to install @user_8211d795/healthcheck-skill

About this skill

Problem

OpenClaw hosts running on VPS, local workstations, containers, or sandboxes face mixed issues: gateway permissions, credential files, plugin supply-chain risks, CVEs, and environment-variable leaks. Manual checks often miss items, especially in restricted sandboxes with limited commands.

How it works

The skill performs environment-aware checks: it starts with read-only commands such as openclaw status, openclaw update status, and openclaw security audit; then computes an A/B/C/D/E security score across CVEs, configuration, malicious skills, prompt injection, sensitive data, and least privilege. For containers, it checks non-root users, read-only mounts, and minimal capabilities. For CVEs, it can query the NVD API, rank severity, and generate alerts. Outputs include dashboards, remediation plans, Markdown/JSON reports, before/after comparisons, and scheduled check scripts.

Boundaries

Changes require user confirmation; in sandboxes and containers it generally recommends configuration-level fixes rather than system service changes. Some commands are marked as examples or environment-dependent. Scoring and alerts depend on local configuration, log directories, and NVD API availability.

Use Cases

  • After deploying an OpenClaw VPS, run security audits, CVE checks, and baseline hardening.
  • Before containerizing OpenClaw, verify non-root user, read-only mounts, and minimal capabilities.
  • When a high-severity CVE arrives, query NVD data and generate OpenClaw remediation guidance.
  • Periodically scan sandboxes for sensitive data exposure and environment-variable leaks.

Best For

  • SREs operating production OpenClaw hosts who need pre-launch baseline checks and hardening.
  • Platform engineers containerizing AI services who must review Docker privileges, mounts, and secret exposure.
  • Skill developers maintaining OpenClaw plugins who need to audit malicious skills, supply-chain risk, and config leaks.
  • Engineers using Coze sandboxes for automation tests who need environment-variable leak and restricted-config security checks.