Back to skill
Skillv1.0.0
ClawScan security
Skill Vet · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
SuspiciousMar 7, 2026, 3:56 AM
- Verdict
- suspicious
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill's code matches its stated purpose (a static vetter) and appears non-malicious, but there are implementation inconsistencies (no declared runtime/binary requirements and no install mechanism for the 'skill-vet' CLI) that should be clarified before trusting it.
- Guidance
- This skill appears to implement a local static scanner and does not exfiltrate data or call external endpoints, but there are practical inconsistencies to resolve before use: - Runtime: vetting.cjs is a Node script. Ensure the agent/runtime has Node available and clarify how to invoke it (the SKILL.md uses the 'skill-vet' CLI name but no install/installation steps are provided). - Scope of scans: The tool reads files under whatever path you give it. Do not run it pointed at directories containing secrets you don't want read, unless you trust the environment. - False positives: Many flagged patterns (Buffer.from, fs.writeFile, console.log, process.env access) can be benign; review findings manually. - Minor bug: the script references colors.gray which is undefined — this is cosmetic but indicates the code wasn't thoroughly tested. If you plan to install or run this skill, ask the publisher for explicit installation instructions (how to install the 'skill-vet' CLI or how OpenClaw will invoke vetting.cjs) and confirm that Node will be available. Otherwise the code itself looks coherent with its stated purpose and not malicious.
Review Dimensions
- Purpose & Capability
- noteThe name/description (a security scanner) align with the included vetting.cjs scanner: it searches files for risky patterns and reports findings. However, the package declares no required binaries or install spec while providing a Node.js CLI script (vetting.cjs). Running this script requires Node on the agent/runtime and a way to invoke it as 'skill-vet' — neither is declared or documented in the SKILL.md, so the runtime expectations are unclear.
- Instruction Scope
- okSKILL.md describes only scanning skill directories and generating reports; the tool's code implements that behavior and does not attempt to read or transmit data off-host. It walks the target path, reads files, and reports regex matches. One note: the README/commands assume a 'skill-vet' executable, but no install instructions are provided.
- Install Mechanism
- concernThere is no install spec. The skill includes a Node .cjs CLI (shebang present) but does not declare Node as a required binary nor provide steps to install the CLI or place it on PATH. This mismatch makes it unclear how the tool is intended to be executed in the target environment.
- Credentials
- okThe skill requests no environment variables, credentials, or config paths. The code scans for patterns like process.env.* in target files (i.e., it detects code that accesses env vars) but does not access the runtime environment's secrets itself.
- Persistence & Privilege
- okalways is false and the code does not persist configuration, modify other skills, or request elevated privileges. The tool only reads files in the target path and exits with non-zero when high-risk findings exist.