Supabase database diff: compare schema and rows before you restore
A database diff answers a different question from a backup: not 'what state can I restore?' but 'what changed between these two states?'.
What a database diff should show
A useful diff separates schema changes from row changes. It should identify tables that were added or removed, columns that changed, and rows that were added, deleted, or modified.
It should also be honest about limits. A table that could not be read or compared should be marked as unavailable instead of being presented as unchanged.
Why compare two Supabase snapshots
Comparing yesterday's snapshot with today's snapshot gives you a focused review surface after a migration, deploy, or unexpected production change. Primary keys make row-level comparison actionable because each change can be tied back to a record.
Snapback uses a read-only connection and stores snapshots separately. The beta keeps up to 7 snapshots for 7 days, so it is designed for recent change review rather than indefinite archival.
A safe review sequence
First compare the snapshots closest to the suspected change. Then filter to the affected table, inspect changed columns and row keys, and only after that choose a recovery or migration action in your normal database workflow.
Snapback's beta does not restore rows. That restriction is intentional: the product shows evidence without silently changing your database.
Snapback