Skill Auditor Pro
PassAudited by ClawScan on May 10, 2026.
Overview
This skill’s shell-based auditing behavior matches its stated purpose, with only minor user-notice items around local scanning, external CLI assumptions, and temporary saved snippets.
This appears reasonable for a security-auditing skill. Before using it, confirm you are running the reviewed scripts/audit.sh, scan only the intended skill directory or skill name, avoid sudo, and remove any /tmp suspicious-code file after analysis.
Findings (3)
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.
If a user points the scanner at a broad local directory instead of a skill folder, it may read more local files than intended.
The script scans all matching files under the user-supplied skill path. This is expected for a skill auditor, but the scope depends on the path the user provides.
find "$skill_path" -type f \( -name "*.md" -o -name "*.sh" -o -name "*.py" -o -name "*.js" \) -print0
Run it only against specific skill directories or skill names you intend to audit.
The audit may fail or behave differently if the expected local tools are unavailable.
The skill documents reliance on local CLI tooling, while the registry requirements list no required binaries. This is a documentation/dependency clarity issue, not hidden installation behavior.
clawhub inspect <skill-name> ... ~/.openclaw/workspace/skills/skill-auditor/scripts/audit.sh <skill-name> ... Requires `gemini` CLI installed
Verify the script path and required CLIs before running the commands.
Potentially sensitive or prompt-injection-bearing snippets from the scanned skill could remain in a temporary file or be shown to an agent.
Suspicious snippets from the scanned skill are saved to /tmp and the user is prompted to have an agent analyze them. This is purpose-aligned, but the content is untrusted and may persist after the scan.
local output_file="/tmp/skill-audit-${skill_name}-suspicious.txt"
cp "$SUSPICIOUS_CODE" "$output_file"
echo -e " ${MAGENTA}分析这段可疑代码: cat $output_file${NC}"Delete the /tmp audit file when finished and explicitly tell any LLM or agent to treat the file contents as untrusted quoted evidence, not instructions.