ClawSkillGuard

PassAudited by VirusTotal on May 11, 2026.

Overview

Type: OpenClaw Skill Name: clawskillguard Version: 1.0.2 The clawskillguard skill is a legitimate security utility designed to perform static analysis on other OpenClaw skills. The scan.py script uses regex patterns to detect common indicators of prompt injection, data exfiltration, and malicious code; these patterns are base64-encoded within the script specifically to prevent the scanner from flagging its own detection strings. The SKILL.md instructions correctly guide the AI agent to perform local security audits without any evidence of hidden malicious intent, unauthorized network activity, or data exfiltration.

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

The scanner may read many installed skill files instead of only one target skill.

Why it was flagged

This directs the agent to inspect multiple local skill directories. That is aligned with a security-audit tool, but it broadens the local files being read.

Skill content
If no path given, offer to scan all installed skills.
Recommendation

Provide an explicit skill path unless you intentionally want a full installed-skill audit.

What this means

The skill relies on local script execution to perform scans.

Why it was flagged

The skill instructs running a local Python script. This is expected for a scanner, but users should understand that installing and using it means executing bundled code locally.

Skill content
python3 <skill_directory>/scripts/scan.py <path_to_skill> [--format text|json] [--severity low|medium|high|critical]
Recommendation

Run it only from the installed skill directory you intended to use, and review results before acting on any install recommendation.

What this means

The documented command may fail or cause confusion about which file should be executed.

Why it was flagged

The documented command references scripts/scan.py, while the supplied manifest lists scan.py at the root and the metadata declares no required binaries. This looks like a packaging/documentation mismatch rather than malicious behavior.

Skill content
python3 <skill_directory>/scripts/scan.py <path_to_skill>
Recommendation

Verify the installed file path before running the scanner; the packaged file appears to be scan.py at the skill root.