DDD Repository Design

A repository gives use cases collection-like access to aggregate roots. It should not become a general database API.

Workflow

1. Identify the aggregate root the repository owns.

2. Keep methods aligned with aggregate retrieval and persistence needs.

3. Put complex read/query screens in query services or read models, not command repositories.

4. Hide ORM, SQL, and transport details behind the implementation.

5. Define optimistic locking, transactions, and not-found behavior explicitly.

Rules

• Repositories return aggregate roots, not internal entities.