Skillv1.0.0

ClawScan security

Skill Doc Formatter · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 1, 2026, 1:17 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code and instructions are coherent with its stated purpose (formatting SKILL.md and running a local security check); it reads and analyzes local skill files but does not request credentials, install remote code, or contact external endpoints.
Guidance: This package appears to do exactly what it says: format SKILL.md files and run a local static security checklist. It's safe to run locally with these caveats: 1) Prefer running without --inplace first (use stdout or -o) so you can review changes before overwriting; 2) The security checker is heuristic (regex-based) and may report false positives/negatives — treat its output as guidance, not authoritative proof; 3) The source/owner is unknown (no homepage), so if you plan to run this on sensitive repositories, review the scripts yourself (they're bundled and small) before use. If you want stronger assurance, inspect scripts/security_review.py and scripts/format_skill_doc.py in full (they are included) or run them in an isolated environment.

Purpose & Capability: okThe name/description (format SKILL.md for ClawHub) matches the provided files and behavior. The repo contains a formatter (scripts/format_skill_doc.py), templates, and a security checker (scripts/security_review.py). No unrelated credentials, binaries, or install steps are requested.
Instruction Scope: okSKILL.md instructs running the formatter against local SKILL.md files and optionally running the bundled security review. The scripts read files in the supplied skill directory, parse frontmatter and sections, and may write output (with --inplace or -o). The security checker inspects files for patterns (subprocess usage, logging, env vars, missing files) but does not send data off-host or execute third-party code. Note: --inplace will overwrite files, so review output before overwriting.
Install Mechanism: okNo install specification is present (instruction-only with shipped scripts). The tool is run with Python from source; it does not download archives or execute installers. This is low-risk and proportionate for a formatter utility.
Credentials: okThe skill declares no required environment variables or credentials. The bundled security checker scans code for uses of env vars and secrets but does not itself require or request credentials.
Persistence & Privilege: okNo persistent/background behavior is requested (always:false). The skill does not modify other skills or global agent settings; it operates on files you point it at.