refactoring

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill contains comprehensive documentation and code snippets intended to teach refactoring patterns and modern code development practices.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: Code examples demonstrate standard patterns for accessing environment variables (e.g., process.env.STRIPE_KEY) and communicating with well-known services like Stripe. These examples follow security best practices and are provided for educational context.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing and refactoring user-provided code. This task inherently possesses a surface for indirect prompt injection if the code being processed contains malicious instructions designed to manipulate the agent.
  • Ingestion points: User-supplied source code provided during refactoring sessions.
  • Boundary markers: Not explicitly defined in the documentation.
  • Capability inventory: The skill is scoped to code analysis, modification, and readability improvements.
  • Sanitization: No specific input sanitization steps are documented for the analyzed code.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 06:57 AM
Security Audit — agent-trust-hub — refactoring