Skillv1.0.0

ClawScan security

Security Monitor V15 T33 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 14, 2026, 6:58 AM

Verdict: Benign
Confidence: medium
Model: gpt-5-mini
Summary: This is a local Linux/Unix security/audit script whose requested capabilities and instructions match its stated purpose, but it requires root to be fully effective and you should review the code before running it on production systems because the source is unknown.
Guidance: This skill appears to be a local host audit tool and is internally coherent, but the source is unknown — review before running. Key points: (1) The script needs access to process lists, network state and filesystem; running as root (recommended for full checks) gives it full read access to sensitive files and secrets — only run on machines you control. (2) Inspect the full scripts/monitor.py for any network calls or hardcoded endpoints (exfiltration risks) before executing; the visible parts use subprocess and local scans which are expected, but the file was truncated in this review. (3) Test in an isolated environment (VM or container) first, and consider restricting scan directories (pass explicit paths) rather than scanning the whole filesystem. (4) If you plan automated runs (cron), ensure alerting integrations (email/webhook) are configured to trusted endpoints and credential handling is safe. (5) Because the skill has no known homepage or publisher reputation, prefer using well-known, audited tools for critical systems or perform a code audit prior to production use.

Review Dimensions

Purpose & Capability: okName/description (server process, network, file-permission, secrets, logs, process-tree auditing) align with included files and runtime behavior. The skill is an instruction-first scanner that reads processes, network state and files — which are appropriate for a host security audit. No unrelated credentials, binaries, or install steps are requested.
Instruction Scope: noteSKILL.md instructs running scripts/monitor.py (optionally as root) and scanning arbitrary directories (/var/log, /opt/openclaw, /home, /etc). That scope is expected for a scanner, but it means the tool will read sensitive files and system state. The README includes an optional SMTP/email snippet as an example (not integrated by default) — review any added alert/integration code before use.
Install Mechanism: okNo install spec — instruction-only plus included Python script. No external downloads or package installs are requested by the skill itself.
Credentials: okNo environment variables, credentials, or config paths are required by the skill metadata. The script requires local filesystem and command-line tools (ps, ss/netstat, find) — which is proportional to a local audit tool.
Persistence & Privilege: notealways:false and no persistent install — the skill does not request permanent inclusion. However SKILL.md recommends running the script as root to obtain full visibility; running as root grants broad read access to system files and secrets, so treat that as a high-privilege operation and only run after review.