effect-sql-db

Transactional SQL with Effect requires wrapping database operations in a transaction service that handles commits, rollbacks, and retry logic. This Skill provides typed transaction patterns for Drizzle, Kysely, Prisma, mysql2, and pg clients, automatically retrying on deadlocks and mapping low-level driver errors to a unified DatabaseError type.

Yes. This Skill integrates Prisma, Drizzle, Kysely, mysql2, and pg with Effect by centralizing database configuration in a service layer, enabling typed error handling, transactional support, and built-in retry policies for deadlock scenarios across all supported clients.

Implement retry policies at the transaction level using Effect's retry combinators. This Skill provides deadlock-aware transactional patterns that automatically retry failed operations according to configurable policies, integrated with your Effect service layer and observable through hooks for monitoring.

This Skill exposes observability hooks for slow query detection and monitoring. Wrap database operations in the provided transaction service to capture execution metrics, integrate with your Effect observability stack, and identify performance bottlenecks across Drizzle, Kysely, Prisma, and native SQL drivers.

Yes, for production Effect applications. A service layer centralizes database configuration, error mapping, and transactional logic, reducing boilerplate and ensuring consistent retry and observability behavior. This Skill provides the service patterns you need for Drizzle, Kysely, Prisma, mysql2, and pg.

This Skill includes streaming examples for query results with Effect. Stream patterns integrate with the transaction and error-handling infrastructure, allowing you to process large datasets efficiently while maintaining retry semantics and observability across Drizzle, Kysely, Prisma, and native drivers.