Skip to content

Core Concepts

pgmt takes a unique approach to PostgreSQL schema management: declarative development with explicit production deployments. These concepts explain why and how.

  1. Philosophy - Why pgmt is built around a schema diffing engine
  2. How pgmt Works - The development workflow from local iteration to production deploy
  3. Shadow Database - How pgmt validates changes without touching your databases
  4. Dependency Tracking - How pgmt ensures operations happen in the right order

Schema files show the current state of your database, not a chain of historical migrations. You edit them directly during development, then generate explicit migrations for production deployment.

You can recreate your database at any time using schema files and migration files. No manual steps required.

Development, staging, and production are completely separate. Shadow databases let you test changes safely before applying them.


Ready to get started? See the Quick Start Guide to try pgmt in practice.