grey-haven-data-modeling

Multi-tenant database design isolates customer data using tenant_id fields on all tables combined with Row-Level Security policies. This Skill automates schema generation for PostgreSQL with RLS enablement, ensuring every row is tagged with a tenant identifier and queries automatically filter by the current tenant, preventing cross-tenant data leaks.

This Skill generates optimized schemas for both SQLModel (Python/FastAPI) and Drizzle ORM (TypeScript) stacks. You can choose your language and ORM; the Skill produces tenant-aware table definitions, migrations, and RLS policies tailored to each framework's conventions and requirements.

PostgreSQL Row-Level Security policies combined with tenant_id columns enforce data isolation at the database layer. This Skill configures RLS with admin override capabilities, applies indexes on tenant_id and foreign keys for performance, and sets up standard audit fields (created_at, updated_at, deleted_at) across all tables.

Multi-tenant relationships require proper foreign key constraints and join tables that inherit the tenant_id context. This Skill automatically generates one-to-many and many-to-many relationship definitions with composite indexes on (tenant_id, foreign_key) to maintain isolation and query performance across related entities.

Multi-tenant tables follow snake_case naming with mandatory fields: id (primary key), tenant_id (isolation), created_at, updated_at, and deleted_at (soft deletes). This Skill enforces these conventions across all generated schemas, ensuring consistency and enabling RLS policies to function correctly.

This Skill generates both the schema structure and the RLS policy configuration required for PostgreSQL. It creates policies that automatically enforce tenant isolation based on tenant_id, with built-in admin override support, so your multi-tenant data protection is active immediately after schema deployment.