keyword-research
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns, prompt injections, or obfuscated content were detected across the skill files.
- [NO_CODE]: The skill contains no executable scripts (Python, JavaScript, or shell). It relies entirely on natural language instructions and structured data processing guidelines.
- [DATA_EXPOSURE]: The skill requests standard, non-sensitive business information (target audience, goals, topics) from the user to perform its analysis. It does not attempt to access system environment variables or sensitive local files.
- [SAFE]: All external references point to legitimate vendor-controlled domains (github.com/openclaw, skills.sh) for documentation and skill management purposes.
- [SAFE]: While the skill involves processing competitor URLs (a theoretical surface for indirect prompt injection), it possesses no elevated capabilities—such as file writing or network requests—that would allow for exploitation.
Audit Metadata