Back to skill
Skillv1.0.0
ClawScan security
AI Code Review · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 8:27 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only code review skill with no code, install steps, credentials, persistence, or broad system access requested.
- Guidance
- This skill appears safe to install as an instruction-only review framework. As with any code review assistant, avoid sharing secrets or proprietary code unless you are comfortable with the agent seeing that content.
Review Dimensions
- Purpose & Capability
- okThe skill's stated purpose is to guide structured code review for security, performance, correctness, and maintainability, and the SKILL.md content is consistent with that purpose.
- Instruction Scope
- okThe instructions focus on reviewing submitted code and producing prioritized feedback; they do not direct the agent to override user intent, make changes, publish content, or take high-impact actions.
- Install Mechanism
- okNo install specification, required binaries, dependencies, or code files are present.
- Credentials
- okThe artifacts request no environment variables, credentials, network access, local indexing, or privileged system capabilities.
- Persistence & Privilege
- okThe artifacts show no persistence, background execution, account access, credential use, or privilege escalation.