create-data-modellisted

# Create Data Model ## Trigger Use when the user asks for entities, database schema, ERD, storage boundaries, data lifecycle, migrations, indexes, row level security, collections, or data ownership — on any supported database. ## When To Use - After the PRD, technical design document, and architecture plan are agreed and handed over. The schema is downstream of those, not a substitute for them. - Before API contracts, and before any persistence-heavy implementation. - When an existing backend has grown without a recorded model and needs one. ## Why This Skill Produces Files, Not Prose An entity list written as Markdown cannot be read back by anything. Later backend work then re-derives the model from whatever code is nearby, and the schema drifts one query at a time. That is the usual cause of a messy backend. So the deliverable is a real schema file. `schema.sql` is the **source of truth**: a human writes and edits it, and it is what ships. Everything else in `data/` is generated from it and must never be hand-edited. On a document store the same role is played by `schema.json`. ## Dialects This skill is not Postgres-only. The dialect decides whether enums are a declared type, whether row level security exists, how identifiers are quoted, and what a migration is allowed to say — so it is resolved **first**, before any design work. Supported: `postgresql` (and Supabase), `mysql` (and MariaDB), `sqlite`, `sqlserver`, `mongodb`. See `references/data-model-adapters.m