code-reviewer
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions explicitly mandate a read-only workflow, prohibiting file modifications, commits, or network-altering actions like pushing branches. It specifically warns against installing dependencies or running package managers that might download untrusted tools. The evaluation cases focus on typical software engineering concerns such as null dereferences, branch comparisons, and adherence to repository-specific rules defined in AGENTS.md. The presence of $code-reviewer in the OpenAI interface configuration is a standard tool invocation pattern for this platform.
Audit Metadata