Skip to content

Database

Schema, migrations, RLS policies, and data management.

Guides

Database Setup

Complete schema reference:

  • Table structure and relationships
  • Postgres stored functions (get_current_snapshot_id, set_current_snapshot, etc.)
  • Row Level Security (RLS) policies
  • Indexes and performance considerations

Data Hydration and Seeding

Comprehensive data strategy reference covering all three layers:

  • Layer 1: Migrations (schema + minimal seed rows)
  • Layer 2: Content seed scripts (curriculum and catalog data)
  • Layer 3: Test user / auth seed (local only)
  • Environment profiles (local, staging, production)
  • All deno task commands for data management

Seed Curriculum Guide

Agentic AI curriculum seed data (the built-in sample dataset):

  • 1 domain · 3 trails · 12 concepts · 18 sparks · 20+ prerequisite links
  • Seed data format
  • Import process

Snapshot Publishing Workflow

Content versioning lifecycle:

  • is_published (per-entity visibility control)
  • snapshot.status (batch version control: draft → published → archived)
  • How both mechanisms work together
  • Historical queries using ?snapshot_id=uuid

Content Model

Domain (e.g., "Machine Learning")
  └── Trail (10-20 hours, e.g., "Transformer Mastery")
       └── Concept (30-60 mins, e.g., "Attention Mechanisms")
            └── Spark (5-10 mins, e.g., "Self-Attention Explained")

Sparks are connected by prerequisite links forming a DAG (knowledge graph). All entities belong to a snapshot enabling versioned queries.

Database Commands

# Reset local database (applies all migrations + seed data)
supabase db reset

# Create new migration
supabase migration new <name>

# Push migrations to remote
supabase db push

# Push to specific branch (staging)
supabase db push --branch develop

# Check migration status
deno task db:status