schema-migration

Detect data-loss risks in SQL schema migrations by comparing old and new DDL to identify type narrowing, missing defaults, dropped constraints, and breaking column changes before applying them to production.

Schema changes that cause data loss in ALTER TABLE statements include irreversible type narrowing, adding NOT NULL columns without a DEFAULT, and dropping columns or constraints during the migration.

Yes, you can check dbt model refactors for breaking schema changes by diffing old and new SQL models or YAML schemas to ensure downstream models remain valid and avoid column-level data loss.

Yes, this schema diff analysis supports YAML and JSON schema files, comparing old and new versions to surface column-level breaking changes and constraint modifications across common SQL dialects.

The best way to prevent irreversible type changes before a production migration is to run a schema diff analysis on old and new DDL to catch narrowing and invalid type conversions early.

You should run a schema diff before applying database changes whenever you execute ALTER TABLE or CREATE TABLE migrations, refactor dbt models, or modify YAML schemas to prevent unexpected data loss.