refactoring
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external code via the code parameter. Ingestion points: SkillParams.code in SKILL.md. Boundary markers: None provided in instructions or schema to delimit user code from agent instructions. Capability inventory: The skill uses ast_parser and smell_detector for analysis, and includes a validation script scripts/validate.py. Sanitization: No sanitization or filtering of the input code is specified. This presents a surface where malicious instructions could be embedded in code comments or strings.
- [SAFE]: No other security concerns were identified.
Audit Metadata