What problem does it solve? AI assistants and developers often forget schema details mid-session, producing code with wrong column names, missing fields, or incorrect types that only fail at runtime. This Skill prevents those errors by mandating schema verification before any database code is written. ## Core Features & Use Cases - Schema-First Rule: Requires reading the schema file (Drizzle, Prisma, Supabase, SQLAlchemy, TypeORM, or raw SQL) and completing a verification checklist before writing any database code. - Type Generation Guidance: Provides commands and patterns for generating TypeScript types from Drizzle, Prisma, and Supabase, plus Pydantic models for SQLAlchemy. - Schema-Aware TDD Workflow: Extends test-driven development with a schema verification step so tests fail on logic, not on schema mismatches. - Use Case: When adding an order history endpoint, the assistant first reads src/db/schema.ts, confirms columns like orders.total_cents exist, maps types to TypeScript, and only then writes type-safe queries. ## Quick Start Ask the assistant to read the project schema file and complete the schema verification checklist before writing any database query or migration.
