django-reviewer
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill follows best practices for a code review tool.
- [COMMAND_EXECUTION]: The skill uses git commands (git diff, git status) to determine the scope of the review. These are standard operations for development agents.
- [PROMPT_INJECTION]: The skill acts on untrusted code which is a surface for indirect prompt injection. However, the instructions include explicit 'Authorization Boundary' rules to prevent the agent from performing unauthorized actions.
- Ingestion points: git diff, git status, local project files, and user-supplied code (SKILL.md).
- Boundary markers: Strong instructions in the 'Authorization Boundary' section to treat requests as report-only by default and never infer write authorization.
- Capability inventory: Reading local files, executing git commands for status and diffs, and potentially modifying files if authorized by the host.
- Sanitization: No explicit code-level sanitization mentioned; the skill relies on the agent following the provided instructions and its own guardrails.
Audit Metadata