domain-model-boundaries-mapper

Bounded contexts are identified by grouping code around distinct domain areas with their own language, entities, and ownership. This Skill maps existing code to domains, documents coupling points, and defines 3-7 contexts with clear boundaries.

An anti-corruption layer is an adapter that translates one bounded context's model into another's, preventing direct coupling. For example, an Order context uses a CustomerAdapter to convert Customer domain data into its own OrderCustomer model.

Shared models are resolved by defining separate models per bounded context, each containing only the fields that context needs. The Skill recommends context-specific interfaces and adapters so changes in one domain do not affect others.

A shared kernel fits when two contexts must share minimal code, such as common value objects like Money or Address. Keep it small; anything larger should be split into separate per-context models to preserve boundary integrity.

DDD context mapping adds structural overhead that may not pay off for small, single-team codebases with one domain. It is most valuable when multiple teams, services, or complex domain logic require explicit ownership and boundaries.