OpenClaw Hardener

AdvisoryAudited by Static analysis on Apr 30, 2026.

Overview

No suspicious patterns detected.

Findings (0)

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

Running the non-read-only modes can change file permissions or OpenClaw runtime configuration.

Why it was flagged

The documented workflows include modes that can apply fixes and change gateway configuration, which are high-impact actions even though they are disclosed and user-directed.

Skill content
python3 skills_live/openclaw-hardener/scripts/hardener.py fix --all ... python3 skills_live/openclaw-hardener/scripts/hardener.py apply-config
Recommendation

Run check --all first, review the results and any config patch, and only then run fix or apply-config if you agree with the changes.

What this means

The skill can inspect and modify important local OpenClaw and workspace files when its commands are run.

Why it was flagged

The skill requests broad read/write authority over the workspace and OpenClaw local configuration area, which may include sensitive local configuration.

Skill content
"read": ["<repo>/**", "~/.openclaw/**"], "write": ["<repo>/**", "~/.openclaw/**"]
Recommendation

Use it only in the intended OpenClaw workspace, keep backups of important configuration, and review output before allowing changes.

What this means

If the workspace audit script is untrusted or has been tampered with, running that path could execute unwanted local code.

Why it was flagged

The helper contains functionality to execute a repo-local security_audit.sh script through bash. This can be appropriate for a hardening workflow, but it means local workspace code may be executed.

Skill content
rc, out, err = run(["bash", str(sh), "--target", str(workspace_root)], timeout_s=900)
Recommendation

Before using all workspace audit features in an untrusted repo, review the referenced security_audit.sh script or restrict use to the built-in OpenClaw audit and read-only checks.

What this means

Users have less external context for verifying the publisher or source history of the skill.

Why it was flagged

The registry metadata does not provide an upstream source or homepage, so provenance is limited even though the included artifacts are coherent.

Skill content
Source: unknown; Homepage: none
Recommendation

Confirm that you trust the publisher and review the included script before installing or running mutation modes.