subagent-review
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data in the form of project source code, which serves as a surface for indirect prompt injection.
- Ingestion points: Files identified in Phase 1: Exploration (SKILL.md) are read by subagents during Phase 2: Review and Phase 3: Verification to identify problems and verify findings.
- Boundary markers: The instructions for subagents in SKILL.md do not specify delimiters or include warnings to ignore instructions that may be embedded within the source code files.
- Capability inventory: Subagents are authorized to write findings and a synthesized fix plan to the local
.review/directory. This plan is designed to be consumed by an orchestration tool for automated changes. - Sanitization: There is no evidence of sanitization or filtering of the source code content before it is processed by the subagents.
Audit Metadata