requesting-code-review
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses git commands (git rev-parse, git diff, git log) to extract change information from the local repository. These are standard development operations for code review workflows.
- [INDIRECT_PROMPT_INJECTION]: The subagent reads content from code changes (git diff) and external plan files, which presents a surface for indirect prompt injection if the codebase contains malicious instructions.
- Ingestion points: Git diff output and plan/requirements documents.
- Boundary markers: Absent; the template does not instruct the agent to ignore instructions embedded in the code under review.
- Capability inventory: The subagent generates a review report, and the parent agent is instructed to act on this feedback, which could lead to executing malicious fixes if the reviewer is compromised.
- Sanitization: No sanitization or validation of the code content is performed before processing.
Audit Metadata