Overview

The CRUD engine is a high-level orchestration layer over Prisma that provides a standardized way to perform database operations across all Prisma models. It ensures consistent handling of cross-cutting concerns like validation, hooks, and observability.

Core Philosophy

• Standardization: Every CRUD operation follows the same execution pipeline
• Extensibility: Hooks can intercept before and after operations
• Type Safety: Full type inference from Prisma models to Hono RPC clients
• Resilience: Built-in retry logic for transient database failures

Key Components

File Structure

features/crud/
├── crud.routes.ts       # Route factory for Hono
├── factory.ts          # Hook registration utilities
├── hooks.ts             # Global hook registry
├── layers.ts            # Effect-TS layer composition
├── services/
│   ├── crud.service.ts         # Main service
│   ├── crud-repository.service.ts  # Prisma wrapper
│   ├── crud-hook.service.ts    # Hook executor
│   └── operation-handler.service.ts  # Operation handler
├── types/
│   └── index.ts         # Context tags and interfaces
└── schemas/
    ├── index.ts         # Schema exports
    └── *.crud.ts        # Per-model CRUD schemas

Operation Pipeline

The CRUD engine uses the shared Operation Pipeline which provides:

1. Prepare Stage — Before hooks execute, validation
2. Execute Stage — Database operation via CrudRepository
3. After Stage — After hooks (success/failure)

Quick Navigation

• Architecture — Request flow and layer composition
• API Reference — createCrudRoutes factory and endpoints
• Services — Effect-TS service implementations
• Observability — Metrics, tracing, and logging
• Runbook — Troubleshooting and maintenance

Supported Operations

Each model exposes these standard operations:

• findOne — Get single record by ID
• findMany — Query multiple records with filters
• createOne — Create new record
• updateOne — Update existing record
• deleteOne — Soft-delete record
• aggregate — Aggregation queries
• groupBy — Grouping queries