Skip to content
BifrostQL
Docs

Ready-Made Database Schemas to Start From

Eight ready-made database schemas ship inside the BifrostQL desktop app. Each one creates a populated SQLite database in a few seconds, so you can see generated queries, joins, and mutations against a realistic shape before pointing BifrostQL at a database of your own. They also serve as reference designs — the CRM schema shows polymorphic notes, the org model shows tenancy.

Every schema is SQLite DDL with foreign keys enforced.

The schemas

Section titled “The schemas”
Schema Tables What it models
blog 6 Authors, posts, categories, comments, and many-to-many tags
classroom 6 Instructors, courses, students, enrollments, assignments, submissions
crm 6 Sales pipeline: companies, contacts, deals, stages, activities, polymorphic notes
ecommerce 7 Catalog, customers, addresses, orders, order items, reviews
project-tracker 8 Workspaces, projects, sections, tasks, labels, multi-assignee assignments
org-model 7 Multi-tenant foundation: tenants, roles, permissions, users, memberships, invitations, audit log
membership-manager 16 Club and association management on the org model: households, members, plans, dues invoices and payments, events, RSVPs, attendance
sqlite-advanced 8 IoT sensor telemetry, and a workout for SQLite features — a generated column and a view

org-model is the foundation membership-manager builds on. Read it first if you are designing a multi-tenant application; the same shape is documented in the multi-tenant organization data model guide.

sqlite-advanced exists to exercise the schema reader. It carries a GENERATED ALWAYS column and a CREATE VIEW, so it is the fastest way to see how BifrostQL surfaces both.

Seed data

Section titled “Seed data”

Every schema installs its DDL. Seed data is separate, and how much you get depends on the schema:

Schema sample seed full seed
blog yes yes
classroom yes yes
crm yes yes
ecommerce yes yes
project-tracker yes yes
membership-manager yes —
org-model yes —
sqlite-advanced yes —

sample seeds are a few hundred rows — enough to read a grid and follow a join. full seeds run to thousands of rows across every table, which is what you want when testing paging, grouping, aggregates, or chart cardinality.

A schema with no seed for the size you asked for installs the DDL and says so. It does not fail.

PostgreSQL variants

Section titled “PostgreSQL variants”

org-model and membership-manager also ship PostgreSQL scripts: org-model-postgres-seed-sample.sql and membership-manager-postgres-seed-sample.sql. These are self-contained — DDL and data in one file — so you run them against a Postgres database yourself with psql, then point BifrostQL at it as an ordinary connection. They are not reachable from the desktop app’s Quick Start.

Creating one

Section titled “Creating one”

The schemas live inside the desktop app as embedded resources. Build it to get a bifrostui binary — see the desktop database explorer guide — or run ./dev-ui.sh from a repository checkout.

Open the app without a connection string and the Quick Start screen offers blog, ecommerce, crm, classroom, and project-tracker as cards. Pick one, pick sample or full, and the app writes a fresh SQLite file to your temp directory, applies the DDL, applies the seed, and connects.

The remaining three — org-model, membership-manager, and sqlite-advanced — have no card yet. They are created through the same endpoint the cards call:

Terminal window
curl -N -X POST http://localhost:5000/api/database/create-quickstart \
  -H 'Content-Type: application/json' \
  -d '{"schema":"membership-manager","dataSize":"sample"}'

The response streams progress as server-sent events. Use the port the desktop app is serving on.

Profiles

Section titled “Profiles”

A schema may ship a profile document that configures BifrostQL beyond the DDL. crm does: crm.bifrost.json declares a showcase profile carrying schema metadata rules — the polymorphic mapping for notes, soft-delete: deleted_at on deals, and hidden timestamp columns — plus the polymorphic and soft-delete modules. The profile is applied when the database is created, so the CRM sample demonstrates those features working rather than merely having the columns for them.

This is the ordinary metadata surface, not a schema-only feature. See cross-cutting modules and transformers for what else you can declare the same way.

Next steps

Section titled “Next steps”