AI Seed Data Generator

The seed script creates records in dependency order — parent tables first, then child tables. Users are created before orders, categories before products, and so on. Foreign key references use actual IDs from previously created records, ensuring referential integrity. The script handles circular dependencies by inserting placeholder records and updating references after.

Idempotent seeds can be run multiple times without creating duplicate data. The script checks for existing records before inserting, uses upsert operations where supported, or truncates tables before reseeding. This means developers can re-run seeds after schema changes or when they want to reset their development database to a known state.

The Faker-generated option uses the Faker library to create realistic names, emails, addresses, phone numbers, product descriptions, prices, dates, and other domain-specific values. The data includes natural variety — different name lengths, mixed case emails, varied order sizes, and realistic date distributions — making your development environment feel like a real application.

Specify the number of records for each entity in your description. The generator creates exactly the amounts you need — whether it is 5 admin users, 100 products, or 1000 orders. Record counts are configurable through variables at the top of the seed file, making it easy to adjust for different environments or testing scenarios.

The Full feature level includes environment-aware seeding — development gets a large, diverse dataset for UI testing, staging gets a moderate dataset for integration testing, and production gets only essential seed data like admin users and configuration records. The script detects the environment and adjusts its behavior accordingly.