Files

T

grabowski b59904fdae Phases 1-5 + rooms/floors, accounts, custom types, users, notifications

Data model
- Properties, rooms (+optional floors), assets (typed custom fields + Zod
  runtime validator + move history), documents (polymorphic scope)
- Projects -> work packages -> tasks -> subtasks
- Decision events (scoped to project/property/asset/work_package)
- Checklist templates + instances, maintenance schedules (time + usage) with
  auto-materialized checklists on event recording
- Wiki (global + per-project) with revisions + tsvector FTS
- Property accounts (utility/meter numbers by kind)
- Notifications table + per-user channel prefs

Infra
- RBAC guards (requireCompany / requireAdmin)
- Storage abstraction: LocalDiskStorage (HMAC signed URLs) + S3Storage
  behind the same interface, switchable via STORAGE_BACKEND
- CSV export for assets / maintenance / decisions
- QR labels: /api/qr SVG endpoint + printable /assets/[id]/label
- Notifications: in-app + SMTP (own server via nodemailer) + Matrix
  (Client-Server API, per-company room) with opt-in per user
- Company switcher + auto-select first company on login

UI
- Topbar: bell with unread count, theme toggle, name, Sign Out (flat)
- Sidebar: main nav + dedicated Admin section (Asset types, Users, Company)
- Nested-route tabs on property / project / asset detail pages
- Admin UIs for users (invite, role, reset pw, deactivate) and company
  settings (default currency, Matrix room id)
- Custom asset type creation + field-def editor with immutable key/type
  guard and auto-deprecate when removing a field still referenced

Graph
- graphify-out/ committed: GRAPH_REPORT.md, graph.html, graph.json

2026-04-23 15:18:11 +07:00

README.md

drizzle migrations

This directory holds SQL migrations generated by drizzle-kit generate from the TypeScript schema in src/lib/server/db/schema/.

Commands:

npm run db:generate — diff the schema vs. the latest snapshot and emit a new .sql file
npm run db:migrate — apply pending migrations to DATABASE_URL
npm run db:push — skip migrations and sync the schema directly (dev only)
npm run db:studio — open the Drizzle Studio UI

After every generate, review the SQL for surprises — especially around:

enum additions (non-blocking) vs. removals (require a separate data migration)
index changes on large tables (use CONCURRENTLY in production rollouts)
anything touching assets.custom_fields — see project_buildfor_life_ops.md memory for the immutable-key policy.