self-review
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is instructed to search for and merge instructions from untrusted project files such as
REVIEW.md,GITLAB-REVIEW.md,AGENTS.md, andCLAUDE.md. If these files are maliciously crafted, they could influence the agent's code review process or force it to inject backdoors during the code improvement phase. - Ingestion points: Files
REVIEW.md,GITLAB-REVIEW.md,AGENTS.md, andCLAUDE.mdin the repository root (referenced inSKILL.md). - Boundary markers: Absent; the content is merged directly into the checklist provided to the sub-agent without delimiters or safety instructions.
- Capability inventory: The agent has the capability to read local files, task sub-agents with specific instructions, and modify source code based on the feedback received.
- Sanitization: No validation, sanitization, or filtering is performed on the content retrieved from the project documentation files.
Audit Metadata