regression-testing
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [External Downloads] (SAFE): The skill references the 'dorny/paths-filter' GitHub Action in its CI/CD integration example. This is a trusted third-party action commonly used in development workflows and is provided as a template for user configuration.
- [Command Execution] (SAFE): Includes standard commands for running test suites ('npm run test:affected', 'npm run test:smoke'). These are expected within the scope of a quality engineering skill and do not involve arbitrary or dangerous command execution.
- [Data Exposure] (SAFE): The skill logic focuses on processing file path changes and test metadata. It does not attempt to access sensitive system files, credentials, or environment variables.
- [Prompt Injection] (SAFE): While the skill ingests external data (bug descriptions and PR metadata), these are typical for the regression testing domain. The impact is limited to the specialized testing context and no markers for safety bypass were found.
Audit Metadata