Back to skill
Skillv1.0.1
ClawScan security
Skill Security Guide · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 6, 2026, 4:12 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This is an instruction-only security guidance document for skill authors; its declared requirements and runtime instructions are consistent with that purpose and do not ask for secrets, installs, or unexpected privileges.
- Guidance
- This guide appears coherent and safe to install, but before using or copying examples: (1) never paste real API keys or secrets into SKILL.md or examples, (2) when implementing the checklist, verify your code matches documentation and does not disable SSL or log secrets, (3) run your own static and dynamic checks on code you produce, and (4) be cautious when converting metadata — the recommended single-line JSON format is a documentation requirement, not a mechanism to hide secrets. If you plan to adapt examples into code, test locally and review any added dependencies or env-vars for necessity before publishing.
Review Dimensions
- Purpose & Capability
- okThe skill is a documentation/guide. It declares no binaries, no env vars, and no install steps — which matches the stated purpose of teaching authors how to pass scans.
- Instruction Scope
- okSKILL.md contains developer-facing best-practice checks and examples (grep commands, code snippets). It does not instruct an agent to read arbitrary files, exfiltrate data, contact unexpected endpoints, or access credentials.
- Install Mechanism
- okNo install spec and no code files. Being instruction-only minimizes disk writes and execution risk.
- Credentials
- okThe skill requests no environment variables or credentials. Example snippets reference env var names only to illustrate best practices and explicitly advise not to print secrets.
- Persistence & Privilege
- okalways is false and the skill does not request persistent or elevated privileges or modify other skills' configs. Normal autonomous invocation is allowed but not unusual for an instruction-only skill.