code-reviewer
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No security issues detected. The skill is designed to assist in code quality and security audits, providing a structured approach for AI agents to evaluate external code. It effectively guides the agent through context understanding, multi-category evaluation, and standardized reporting.\n- [PROMPT_INJECTION]: The skill has an inherent surface for indirect prompt injection as it processes user-provided code. This is consistent with its primary function as a code review tool.\n
- Ingestion points: Code or pull requests provided by users as input (documented in
SKILL.md).\n - Boundary markers: Not explicitly defined in the instructions for segmenting untrusted code from instructions.\n
- Capability inventory: The skill references reading surrounding files to gain context (
SKILL.md).\n - Sanitization: No explicit sanitization or filtering of the provided code is performed prior to analysis.
Audit Metadata