Back to skill
Skillv1.0.0
ClawScan security
Product Description Writer · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 3, 2026, 1:22 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only copywriting skill that does not request credentials, install code, run commands, or persist data.
- Guidance
- This skill appears safe to install from the provided artifacts. As with any marketing-copy tool, users should verify generated claims for accuracy and compliance with marketplace rules before publishing.
Review Dimensions
- Purpose & Capability
- okThe stated purpose is to generate ecommerce product descriptions, and the SKILL.md content stays within that copywriting and SEO-writing scope.
- Instruction Scope
- okThe instructions define user-provided inputs and text outputs; they do not direct the agent to override user intent, access hidden data, or take autonomous external actions.
- Install Mechanism
- okThere is no install specification and no code files, so there is no package, script, binary, or dependency execution to review.
- Credentials
- okNo environment variables, binaries, local files, network access, or credentials are requested, which is proportionate for an instruction-only writing skill.
- Persistence & Privilege
- okThe artifacts show no persistence, background activity, credential use, privileged access, or mutation authority.