Back to skill
Skillv1.0.0
ClawScan security
coding-as-dressing · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 5:26 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only teaching skill that explains code using clothing metaphors and does not request tools, credentials, installation, persistence, or system access.
- Guidance
- This skill appears safe to install as a writing/explanation aid. It may suggest code cleanup conceptually, but the artifacts do not instruct it to edit files or run commands automatically.
Review Dimensions
- Purpose & Capability
- okThe stated purpose and SKILL.md content are coherent: it provides a metaphor-based explanation style for helping beginners understand code.
- Instruction Scope
- okInstructions are limited to explanation style, timing, examples, and output format; they do not ask the agent to override user intent, run tools, or perform autonomous actions.
- Install Mechanism
- okThere is no install spec, no required binaries, no dependencies, and no code files; the artifact is instruction-only.
- Credentials
- okThe skill does not request file, network, shell, credential, account, or environment access.
- Persistence & Privilege
- okNo persistence, background behavior, privilege escalation, credential use, or memory/storage behavior is described.