data-architecture

Use when an approved system design exists and the team needs production-grade operational data architecture before engine-specific implementation. Produces data ownership boundaries, engine selection, consistency model, schema strategy, indexing posture, partitioning and sharding rules, replication topology, cache architecture, retention and deletion policy, migration strategy, and operational handoff guidance. Do not use for analytics pipelines, warehouse modeling, stream processing, single-table tweaks, ORM code, or engine-specific DDL; use the relevant implementations/data/<engine> skill instead.