Claude Code Guide
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill is entirely documentation-focused and provides helpful context for developers.
- [PROMPT_INJECTION]: The skill mentions 'forget previous instructions' as a troubleshooting method for users to reset their agent's state. In this context, it is a legitimate usage tip for manual context management rather than a malicious attempt to bypass safety guidelines.
- [COMMAND_EXECUTION]: The guide includes standard developer command examples such as 'npm run dev' and 'npm test' within configuration templates. These are routine development actions and do not represent a security threat.
Audit Metadata