Back to skill
Skillv1.0.0
ClawScan security
Mental Models · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 6:59 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only reasoning aid that applies mental models to user questions and does not request tools, credentials, installation, persistence, or external access.
- Guidance
- This skill appears safe to install as a reasoning framework. Its main effect is stylistic: it may encourage the agent to analyze problems through multiple mental-model lenses.
Review Dimensions
- Purpose & Capability
- okThe stated purpose is to analyze problems using a multidisciplinary mental-model framework, and the artifact content matches that purpose.
- Instruction Scope
- okThe instructions guide analytical framing only; they do not redirect user goals, override higher-priority instructions, require tool use, or ask the agent to take actions outside discussion.
- Install Mechanism
- okNo install spec, binaries, code files, dependencies, or setup commands are present.
- Credentials
- okThe skill does not request environment access, local files, network access, credentials, or system permissions.
- Persistence & Privilege
- okNo persistence, background behavior, account privileges, credential use, or long-running activity is described.