Quality Assurance

The project QA targets the following areas: code style, static analysis, testing, and code coverage.

Code Style & Static Analysis

Each runtime has dedicated tooling for formatting and static analysis, all enforced automatically in both the pre-commit hook and the CI pipeline.

• Kotlin: formatted with ktfmt and statically analyzed with Detekt, applied to appointments-service and documents-service.
• Node.js: formatted with Prettier and linted with ESLint, applied to users-service, master-data-service, and frontend-service.
• Python: formatted and linted with Ruff in ai-service.

Testing

Unit Testing

Kotlin services use Kotest as the testing framework, including its built-in mocking utilities for isolating dependencies. Node.js services use Jest for unit tests and mocking.

Integration Testing

All microservices use Testcontainers for integration tests, which spins up real, throwaway instances of external dependencies (databases, message brokers, etc.) for the duration of the test run, ensuring that each service is tested against a realistic environment without requiring a permanently running infrastructure.

Code Coverage

Coverage is collected for all services and uploaded to Codecov. Kotlin services use Kover, with infrastructure layer packages excluded from metrics as they are better validated by integration tests. Node.js services use the built-in Jest coverage reporter. The full report is available at Codecov.

Quality Control

Quality control is enforced at two levels: locally before each commit, and in the CI pipeline on every push and pull request.

Husky runs lint-staged on every commit attempt, processing only the staged files to keep the hook fast. JavaScript, TypeScript, JSON, and CSS files are automatically formatted with Prettier, while Kotlin files are processed by a custom script (scripts/ktfmt-staged.sh) that invokes ktfmt on staged .kt files.

The full quality control suite runs in the CI pipeline, described in detail in the next section.