Key Concepts

Before getting started, here are the terms you'll encounter in Twins.

Clone

A clone is a one-time operation that copies components from one Supabase project (source) to another (target). Each clone is recorded as a job with a status (pending, in progress, completed, failed) and a detailed report.

A clone can include all or some of the components: schema, data, functions, storage, etc. You choose what gets copied via the clone options.

Twin (pair)

A twin is a persistent pair between a source project and a target project. When you run a clone between two projects, Twins automatically creates a twin.

Twins are used to:

  • Quickly find your project pairs on the dashboard
  • Re-run a clone without re-selecting the projects
  • View the history of all clones performed between these two projects
  • Configure ignore rules specific to this pair
  • Schedule automatic clones for this pair

Component

A component is a type of clonable object in a Supabase project. Twins handles 13 components:

ComponentExample
Tables (schema)CREATE TABLE users (...)
DataThe rows in your tables
ViewsCREATE VIEW active_users AS ...
FunctionsCREATE FUNCTION calculate_total(...)
TriggersCREATE TRIGGER on_user_created ...
RLS PoliciesCREATE POLICY "users can read own data" ...
Edge FunctionsServerless functions deployed on Supabase
StorageBuckets and files (images, documents, etc.)
Auth UsersSupabase Auth user accounts
Cron JobsScheduled tasks via pg_cron
WebhooksHTTP notifications on DB events
Realtime ConfigWhich tables are published in real time
Project ConfigOAuth providers, API settings

Each component can be enabled or disabled individually during a clone.

Connected account

A connected account is a Supabase account linked to Twins. You can connect one or more Supabase accounts via:

  • OAuth — direct connection with your Supabase account (recommended)
  • Personal Access Token (PAT) — a token generated from Supabase settings

Each connected account grants access to all projects and organizations associated with that Supabase account.

Preview

The preview is the step before launching a clone. It retrieves the source project's metadata to show you:

  • The number of tables, rows, functions, etc.
  • Warnings (large tables, missing extensions, etc.)
  • An overview of what will be cloned

Verification

Verification is a post-clone check. It compares the source project and the target project to detect:

  • Missing tables or columns
  • Row count differences
  • Objects that were not cloned correctly

Scheduled clone

A scheduled clone is a clone that runs automatically at regular intervals (every 2h, 6h, or 8h). It is associated with a twin and reuses the same clone options for each execution.

Useful for keeping a staging environment synchronized with production.

Ignore rules

Ignore rules allow you to exclude certain tables from data cloning. They are configured at the twin level and apply to all clones (manual and scheduled) for that pair.

Example: exclude the analytics_events table that contains millions of rows unnecessary for staging.

Backup

A backup is a snapshot of cloned data, saved to S3. If enabled during the clone, it allows you to restore the data later without needing the source project.

Next steps