adversarial-review
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of markdown instructions for the AI agent to perform code reviews. It does not include any scripts, external dependencies, or commands that interact with the system or network.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted code (diffs and files), representing a surface for indirect prompt injection. However, it defines no dangerous capabilities to be exploited.
- Ingestion points: The skill ingests untrusted code via the 'Scope' input (diffs, branches, or files) described in SKILL.md.
- Boundary markers: None explicitly defined for separating input code from instructions.
- Capability inventory: No scripts or tool invocations are defined within the skill files.
- Sanitization: No specific sanitization or filtering of input code is mentioned.
Audit Metadata