code-review-and-quality
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a disciplined code review workflow across five distinct axes: Correctness, Security, Maintainability, Architecture, and Performance. It prioritizes evidence-based reporting using specific file and line citations.
- [SAFE]: The agent's autonomy is appropriately restricted; the instructions explicitly state that a review verdict is not permission to merge or publish, and the agent should not edit code unless repairs are specifically requested.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and process untrusted data (code diffs, PRs, and repository files).
- Ingestion points: Local staged/unstaged changes, PR/commit diffs, and repository specification files as defined in
SKILL.md. - Boundary markers: The skill does not explicitly define unique delimiters or 'ignore' instructions for the content it reviews, relying on the agent's internal context handling.
- Capability inventory: The skill can read local files, write file modifications (if fixes are requested), and execute test suites as mentioned in the workflow.
- Sanitization: There are no explicit instructions for sanitizing or escaping the content of the reviewed code before processing.
- Assessment: The risk is classified as low because the skill's primary function is diagnostic and analytical, and its capabilities are scoped to the development environment.
Audit Metadata