@fayz-ai/db

The shared Drizzle spine that every Fayz plugin schema composes with.

Status: beta — published to npm and used across Fayz dogfood apps. Pre-1.0: minor APIs may change before 1.0.

When apps are composed from plugins, their data models have to compose too. @fayz-ai/db is the schema spine: a small set of canonical Drizzle tables — tenants, people, orders, appointments, products — plus the column helpers (tenant id, timestamps) that plugin schemas build on. Every plugin references the same spine, so a CRM's clients and an agenda's appointments agree on what a person and a tenant are.

It also re-exports drizzle-orm/pg-core so the whole stack runs on one Drizzle instance — apps compose their own tables, the spine refs, and plugin schemas without the dual-copy PgColumn type clashes you get from mismatched drizzle-orm versions.

What's inside

• Spine tables — tenants, people, orders, appointments, products, orderItems (Ring 0 references plugins point at, in public)
• Column helpers — tenantId, timestamps, createdAt
• Re-exported pg-core — the full drizzle-orm/pg-core builder surface, so every package shares one drizzle-orm instance

Install

npm install @fayz-ai/db

Depends on drizzle-orm. Import pg-core builders from here, not from drizzle-orm directly.

Usage

import { pgTable, text, tenantId, timestamps, people } from '@fayz-ai/db'

export const notes = pgTable('notes', {
  id: text('id').primaryKey(),
  personId: text('person_id').references(() => people.id),
  body: text('body'),
  tenantId: tenantId(),
  ...timestamps,
})

Part of the Fayz SDK

The data spine beneath every plugin schema; apps compose it in their own drizzle.config.

Roadmap & contributing

Built and evolving in the open. See the Fayz SDK roadmap for current gaps, missing features, and good first issues.