code-review
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes git commands (
git diff,git log,git rev-parse) to retrieve changes from the local repository for review. This is expected functionality for a code review tool. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository, creating a surface for indirect prompt injection where malicious content in code or documentation could attempt to influence the sub-agents' analysis.
- Ingestion points: Code diffs, commit logs, and documentation files (such as
CODING_STANDARDS.md,CONTRIBUTING.md, and spec files) are collected in Steps 1, 2, and 3 ofSKILL.mdand included in sub-agent prompts. - Boundary markers: The sub-agent prompts in Step 4 do not utilize distinct boundary markers or delimiters to isolate the ingested data from the analysis instructions.
- Capability inventory: The sub-agents analyze the text and report findings; the skill does not grant these sub-agents capabilities to execute commands or perform network operations based on the data.
- Sanitization: No sanitization or filtering is applied to the repository content before it is interpolated into the prompts for the sub-agents.
Audit Metadata