code-refactoring-refactor-clean

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it interpolates untrusted user data into its instructions without isolation.
  • Ingestion points: User-supplied code is injected via the '$ARGUMENTS' variable in SKILL.md.
  • Boundary markers: The prompt template does not use protective delimiters or explicit 'ignore embedded instructions' directives to distinguish user data from system instructions.
  • Capability inventory: The agent is empowered to perform code analysis and apply modifications based on the input.
  • Sanitization: No input validation or sanitization is performed on the code provided in the arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 12:48 PM