Back to skill
Skillv1.0.0
ClawScan security
品牌 GEO 降维提炼师 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMay 1, 2026, 8:24 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5.5
- Summary
- This is an instruction-only brand-copy rewriting skill with no code, tools, credentials, installation steps, or persistence shown in the artifacts.
- Guidance
- This appears safe to install as a text-formatting and marketing-analysis skill. Users should still review the generated marketing advice critically, especially because the skill promotes a specific GEO/brand-ecology framework, but the artifacts do not show security-relevant behavior.
Review Dimensions
- Purpose & Capability
- okThe artifacts consistently describe a writing/transformation skill that converts brand promotional text into a structured Chinese GEO report using the stated tree/root and buyer-intent framework.
- Instruction Scope
- okThe instructions constrain output format, language, and classification style; they do not ask the agent to access external systems, override user goals beyond the skill task, or perform hidden actions.
- Install Mechanism
- okNo install spec, binaries, dependencies, scripts, or code files are present; the skill is instruction-only.
- Credentials
- okThe metadata declares no required environment variables, credentials, binaries, config paths, OS access, or capability tags, which is proportionate for a text-processing skill.
- Persistence & Privilege
- okThe artifacts show no persistence, background execution, account access, memory storage, or privileged local/system behavior.