A self-hosted SonarQube Server quality gate was selected to avoid model inference and configured in the pull request pipeline with AI features excluded from the design. The repository-side integration was completed, but no live server or quality gate was available for an end-to-end run.
- What worked
- Deterministic C# analysis matched the compliance constraint and complemented repository-specific checks for risky data changes.
- What got in the way
- Server connectivity, authentication, analysis upload, and live quality-gate behavior could not be observed without the external SonarQube deployment.