Skip to content

README.md

Latest commit

 

History

History
55 lines (34 loc) · 3.42 KB

README.md

File metadata and controls

55 lines (34 loc) · 3.42 KB

Database Schema & Data Integrity (packages/db)

The @repo/db package is the foundational data layer of the Respondo ecosystem. It centralizes our PostgreSQL schema definitions and the Prisma ORM configuration, ensuring that every application and package within the monorepo interacts with a single, strictly typed source of truth. By decoupling the database logic from the application code, we maintain high data integrity and simplify the orchestration of our relational models.

🏗 Package Structure & Exports

This package is designed to be a "plug-and-play" dependency for any service requiring database access:

  • . (Root Export): Provides the fully initialized and configured Prisma client instance.
  • ../generated/prisma/client: Exports the auto-generated TypeScript types and enums, ensuring that our database queries are end-to-end type-safe.
  • src/client.ts: Handles the instantiation logic. It uses the PostgreSQL adapter and implements a singleton pattern in development to prevent exhausting database connection limits during hot-reloads.
  • prisma/schema.prisma: The heart of the package—the declarative source of truth for our entire database structure.

🗄 Core Data Models

Currently, the schema is optimized to support our robust authentication and session management systems, integrated directly with Better Auth:

  • User: Stores core identity information and profile metadata.
  • Session: Manages active user sessions across devices.
  • Account: Bridges users to various social providers (GitHub, Google).
  • Verification: Handles secure tokens for email verification and credential management.

As Respondo evolves, this schema will expand to house the high-density data required for multi-channel ticketing, customer profiles, and AI-retrieval (RAG) embeddings.

⚙️ Technical Implementation

  • High-Performance Adapter: We utilize @prisma/adapter-pg paired with the pg driver for reliable, native communication with our PostgreSQL instance.
  • Isolated Generation: To avoid conflicts within the monorepo, Prisma generates its output into a local generated/ directory within this package, which is then re-exported for consumer use.
  • Environment Requirement: The package requires a valid DATABASE_URL to establish a connection to the PostgreSQL cluster.

🛡 Maintaining Data Integrity

At Respondo, we treat our schema as a living document that requires careful stewardship:

  1. Strict Typing: Always use the generated types from @repo/db to prevent runtime errors and data drift.
  2. Centralized Migrations: All changes to the database structure must originate in the schema.prisma file here, ensuring that migrations are version-controlled and predictable.
  3. Scalability: The relational structure is designed to support the complex JOIN operations required for advanced analytics and customer history tracking.

🚀 Future Expansion

The next phase of development for this package includes the introduction of models for:

  • Tickets & Conversations: Managing the lifecycle of customer queries.
  • Organizations & Memberships: Supporting multi-tenant environments and team RBAC.
  • Knowledge Base Embeddings: Storing the vectorized data required for our RAG-powered AI engine.

When extending the schema, always remember to run the generation scripts to keep the client types in sync with the underlying tables.