Empty database?
Not for long.

Donk reads your column names and generates realistic data. A column called email gets emails. A column called price gets numbers. No fixtures. No config.

You ran the migration. The tables exist. They're empty. You need data to build against — a list of users, some products, a few orders. Writing seed files is tedious. Copying test data from production is a security risk. Typing test into every field produces data that looks fake because it is.

Donk's Smart Seeder reads your schema — column names and types — and generates contextually appropriate data. A column called first_name gets a real first name. A column called email gets a valid email. A column called price gets a decimal number in a reasonable range. A column called created_at gets a timestamp in the last 90 days.

  • Reads column names and types to infer the right data generator
  • 50+ column name patterns recognized — name, email, phone, url, price, status, etc.
  • Foreign key awareness — generates valid references between tables
  • Configurable row count — 10, 50, 100, or custom
  • Works across all five database engines
  • One click from the database viewer — "Seed this table"
  • CLI access — donk db seed mydb --table users --rows 100
🎬

Video: 30–40 seconds. Empty users table → click Seed → 50 rows of realistic data appear. Show the column name → data type mapping.

Download Donk

Windows 10/11 · ~100 MB · No dependencies

Download for Windows