# NestJS Database

> Data access patterns, Scaling, Migrations, and ORM selection.

- Skill: `majiayu000/nestjs-database` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add majiayu000/nestjs-database`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/nestjs-database/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/nestjs-database

---


# NestJS Database Standards

## **Priority: P0 (FOUNDATIONAL)**

Database integration patterns and ORM standards for NestJS applications.

## Selection Strategy

See [references/persistence_strategy.md](references/persistence_strategy.md) for database selection matrix and scaling patterns (Connection Pooling, Sharding).

## Patterns

- **Repository Pattern**: Isolate database logic.
  - **TypeORM**: Inject `@InjectRepository(Entity)`.
  - **Prisma**: Create a comprehensive `PrismaService`.
- **Abstraction**: Services should call Repositories, not raw SQL queries.

## Configuration (TypeORM)

- **Async Loading**: Always use `TypeOrmModule.forRootAsync` to load secrets from `ConfigService`.
- **Sync**: Set `synchronize: false` in production; use migrations instead.

## Migrations

- **Never** use `synchronize: true` in production.
- **Execution**: Run via init container or CD step.
- **Zero-Downtime**: Use Expand-Contract pattern (Add -> Backfill -> Drop).
- **Seeding**: Use factories for dev data; only static dicts for prod.

## Best Practices

1. **Pagination**: Mandatory. Use limit/offset or cursor-based pagination.
2. **Indexing**: Define indexes in code (decorators/schema) for frequently filtered columns (`where`, `order by`).
3. **Transactions**: Use `QueryRunner` (TypeORM) or `$transaction` (Prisma) for all multi-step mutations to ensure atomicity.