data-model-review
Data Model Review
A first-principles review and redesign skill for data models. Reviews what exists, catalogs every problem, then uses Go's type system to design a model where invalid states are unrepresentable.
Why This Matters
Most data models grow organically. They start as a few structs, accumulate pointer fields, sprout string-typed enums, and eventually become a minefield of nil-pointer dereferences and runtime validation masquerading as type-safe code. This skill stops that rot and replaces it with intentional design.
The output is not just a list of problems. It is a complete redesign — using branded types, interface-based unions, generics, composition via embedding and interfaces, and zero-value discipline — delivered as a self-contained HTML presentation that the team can read, bookmark, and reference.
Process
READ, UNDERSTAND, RESEARCH, REFLECT before each action.