Skip to main content
Agent skill

Healthcheck

Audit/harden OpenClaw hosts: SSH, firewall, updates, exposure, backups, disk encryption, gateway security.

Get started

Get the complete skill folder, including its scripts and reference files.

When to use this skill

Use Healthcheck when an AI agent needs a reusable SKILL.md workflow for this job: Audit/harden OpenClaw hosts: SSH, firewall, updates, exposure, backups, disk encryption, gateway security.

When not to use it

Skip Healthcheck when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.

How to install

Where will you use it?

Install scope
Claude Code folder setup
mkdir -p ~/.claude/skills/healthcheck
# Save SKILL.md to ~/.claude/skills/healthcheck/SKILL.md
  1. Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/healthcheck/SKILL.md. For a shared project, commit the folder at .claude/skills/healthcheck/ instead.
  2. Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
  3. For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.

Full install guide for Claude, Cursor, and Codex →

Try a starter prompt

After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.

Starter prompt
Use the Healthcheck skill for this task: Use Healthcheck when this documented workflow matches the task. Confirm any required inputs or files before making changes.

What supports this recommendation?

68/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.

Documentation
70/100
Practical detail
68/100
Source evidence
63/100
Source signals
70/100
Model compatibility evidence
  • claudeuntested

    No model-specific signal or recorded compatibility test was found.

  • chatgptuntested

    No model-specific signal or recorded compatibility test was found.

  • geminiuntested

    No model-specific signal or recorded compatibility test was found.

  • copilotuntested

    No model-specific signal or recorded compatibility test was found.

  • llamauntested

    No model-specific signal or recorded compatibility test was found.

  • perplexityuntested

    No model-specific signal or recorded compatibility test was found.

  • mistraluntested

    No model-specific signal or recorded compatibility test was found.

  • grokuntested

    No model-specific signal or recorded compatibility test was found.

Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.

What this skill does

# OpenClaw host healthcheck

Goal: assess host risk, run read-only checks, then propose staged hardening without breaking access.

## Rules

- Ask before state-changing actions. - Do not change SSH/firewall/remote access until access path is confirmed. - Prefer reversible steps and rollback notes. - Never claim OpenClaw manages OS firewall, SSH, or updates. - If identity/role unknown, recommend only. - User choices: numbered list. - Never print secrets.

## Context to infer first

- OS/version, container vs host. - Privilege level. - Access path: local, SSH, RDP, tailnet. - Network exposure: public IP, reverse proxy, tunnel, LAN only. - OpenClaw gateway status, bind, auth. - Backup status. - Disk encryption. - Automatic security updates. - Usage mode: personal workstation, local assistant box, remote server, other.

Ask only for missing facts. Simple phrasing preferred.

## Read-only checks

Ask once for permission to run read-only checks. Then run relevant commands.

Common:

```bash openclaw security audit --deep openclaw gateway status --deep openclaw doctor --lint ```

`doctor --lint` can exit `1` for findings: read the report and continue the remaining checks. Ordinary `doctor` and `doctor --non-interactive` can copy legacy config and migrate state without `--fix`; reserve repairs for explicit approval. Read-only checks exclude config/service repairs and state migrations, but may produce incidental logs or cache bookkeeping.

macOS:

```bash sw_vers lsof -nP -iTCP -sTCP:LISTEN /usr/libexec/ApplicationFirewall/socketfilterfw --getglobalstate pfctl -s info tmutil status fdesetup status softwareupdate --schedule ```

Linux:

```bash cat /etc/os-release ss -ltnup || ss -ltnp ufw status || firewall-cmd --state || nft list ruleset systemctl status ssh sshd lsblk -f ```

Windows:

```powershell systeminfo Get-NetFirewallProfile Get-BitLockerVolume ```

## Risk profile

After context is known, ask desired posture:

1. Convenience: local/private, minimal prompts. 2. Balanced: secure defaults, low friction. 3. Strict: remote/public/sensitive data, more lock-down.

## Report shape

- Current posture: one paragraph. - Findings: severity + evidence + why it matters. - Recommended plan: staged, reversible. - Commands: read-only first; write actions only after approval. - Gaps: what could not be checked.

## Hardening menu

Offer only relevant items:

- Bind gateway to loopback/LAN/tailnet intentionally. - Require auth for remote access. - Close public ports or restrict by firewall. - Enable OS security updates. - Enable disk encryption. - Verify backups and restore path. - Disable password SSH or require keys/MFA where appropriate. - Add scheduled `openclaw security audit --deep`.

Confirm exact action before applying.

Intended uses

  • Use Healthcheck when this documented workflow matches the task.

Implementation notes

  • Review the source instructions and adapt inputs before running the workflow.

Try it, then tell us how it went

A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.