android-clean-architecture

Android clean architecture prevents domain leakage by enforcing strict module boundaries where the domain layer remains pure Kotlin and never depends on data, presentation, or frameworks. This keeps business logic framework-agnostic and maintainable.

The best way to implement the repository pattern in Kotlin Multiplatform is by defining repository interfaces in the domain layer and implementing them in the data layer. This approach coordinates local and remote data flows using SQLDelight and Ktor while mapping results into pure Kotlin domain models.

You use Room and Ktor in the data layer to coordinate local and remote data flows. The data layer implements repository interfaces defined in the domain layer, using mappers to translate between database entities, network DTOs, and pure Kotlin domain models.

Clean architecture in Android requires dependency injection to wire UseCases and repository implementations together. The skill provides patterns for DI wiring, ensuring that UseCases remain pure Kotlin and receive their repository dependencies through inversion of control.

To handle errors in UseCases without leaking framework dependencies, you use pure Kotlin constructs like Result or Try wrappers. This ensures the domain layer remains framework-agnostic while still propagating and managing failures from repository operations cleanly.

Yes, you can apply clean architecture boundaries to both Android and Kotlin Multiplatform modules. The skill defines a unified structure that uses Room for Android local storage and SQLDelight for KMP, ensuring domain logic stays shared and framework-agnostic across platforms.