Openclaw Guardian

Security checks across static analysis, malware telemetry, and agentic risk

Overview

Review recommended: it sets up an always-running watchdog that can automatically repair and roll back your OpenClaw workspace, but the script it tells you to run is not included for review.

Before installing, obtain and review the missing `guardian.sh` and referenced setup documentation, confirm exactly what files it can modify, back up `~/.openclaw/workspace`, and decide whether unattended repair and `git reset --hard` rollback are acceptable for your OpenClaw instance.

Static analysis

No static analysis findings were reported for this release.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Risk analysis

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

If the watchdog misdetects a failure or chooses the wrong rollback point, it could overwrite or discard recent OpenClaw workspace changes.

Why it was flagged

The skill describes unattended use of repair, restart, hard-reset rollback, and snapshot commands against the OpenClaw workspace. These are purpose-aligned for a watchdog, but they can modify or discard workspace state without per-action user approval.

Skill content
Detect Gateway down (every 30s) ... Run `openclaw doctor --fix` ... `git reset --hard` to last stable commit, restart Gateway ... Daily automatic git snapshot of workspace
Recommendation

Inspect the actual watchdog script before use, keep separate backups, test in a non-production workspace, and only enable automatic rollback if you are comfortable with unattended `git reset --hard` behavior.

What this means

The most important code path is unavailable for review, so users cannot verify what the background process will actually do.

Why it was flagged

The provided file manifest contains only SKILL.md and no `scripts/guardian.sh`, even though that missing helper is the executable watchdog the user is told to install and run.

Skill content
Copy `scripts/guardian.sh` from this skill to `~/.openclaw/guardian.sh`
Recommendation

Do not run the skill until the referenced `guardian.sh` and setup documentation are included, reviewed, and match the described behavior.

What this means

The process can keep running and taking automated actions until stopped or removed from startup.

Why it was flagged

The skill clearly discloses a long-running background watchdog. That persistence is expected for the stated purpose, but users should understand it will continue operating after setup.

Skill content
Guardian is a standalone bash watchdog that keeps OpenClaw Gateway alive 24/7 ... nohup ~/.openclaw/guardian.sh >> /tmp/openclaw-guardian.log 2>&1 &
Recommendation

Make sure you know how to stop the process, remove the startup entry, and review its logs before enabling it on important systems.

What this means

Anyone who obtains the webhook URL may be able to post to the configured Discord channel.

Why it was flagged

A Discord webhook URL is a delegated posting secret. The alerting use is optional and purpose-aligned, but it is not declared in metadata and the instructions suggest persisting it in a startup script.

Skill content
export DISCORD_WEBHOOK_URL="https://discord.com/api/webhooks/..." ... Or add to `start-gateway.sh` as a persistent export.
Recommendation

Use a dedicated low-privilege webhook, store it carefully, avoid logging it, and rotate it if it may have been exposed.