Workflows
Blocks
Database
Supabase

Supabase

The Supabase block connects your workflow to a Supabase project. It covers the whole platform: table rows, user auth, file storage, Postgres RPC functions, Edge Functions, and pgvector similarity search — 14 operations in one block.

One credential, whole project: Connect once with your project URL and API key, then pick the operation you need from a single dropdown.

⚠️

Looking to start a workflow when a Supabase table changes? Use the separate Supabase Trigger instead. This page covers the action block that reads and writes data mid-flow.

Key Features

  • Row Operations: Insert, get, update, delete, and list rows via PostgREST
  • Auth Management: Create users, send invites, and generate magic links
  • Storage: Upload files, create signed URLs, and list objects in buckets
  • Functions: Call Postgres RPC functions and invoke Edge Functions
  • Vector Search: Run pgvector similarity search through your match function

Credentials

Click the credentials dropdown and create a Supabase Project connection:

FieldDescription
Project URLYour Supabase project URL, e.g. https://xyzcompany.supabase.co
API KeyA service role key or an anon key
⚠️

The Auth operations (Create User, Invite User, Generate Magic Link) require the service role key. The anon key is enough for row reads that your Row Level Security policies allow.


Operations

OperationWhat it doesMain fields
Insert RowAdds a row to a tableTable, Row Data (JSON)
Get RowFetches one row by a column matchTable, Match Column, Match Value, Select Columns
Update RowUpdates matching rowsTable, Match Column, Match Value, Row Data (JSON)
Delete RowDeletes matching rowsTable, Match Column, Match Value
List RowsLists rows with filter, order, and limitTable, Select Columns, Filter Column, Filter Value, Order By, Limit
Auth: Create UserCreates a user accountEmail, Password, Email Confirmed, User Metadata (JSON)
Auth: Invite UserSends an invite emailEmail, Redirect To, User Metadata (JSON)
Auth: Generate Magic LinkCreates a passwordless sign-in link (returns action_link)Email, Redirect To
Storage: Upload FileUploads a file from a URL or from raw textBucket, Object Path, File URL or File Content, Content Type, Upsert
Storage: Get Signed URLCreates a temporary download linkBucket, Object Path, Expires In (seconds, default 3600)
Storage: List ObjectsLists files in a bucketBucket, Prefix, Limit
Call RPC FunctionCalls a Postgres function with JSON paramsFunction Name, Params (JSON)
Invoke Edge FunctionCalls functions/v1/{name} with a JSON bodyFunction Name, Body (JSON)
Vector Search (pgvector)Runs a similarity search through your match functionMatch Function (default match_documents), Query Embedding, Match Threshold, Match Count, Filter (JSON)

Vector Search calls a Postgres function in your project with query_embedding, match_threshold, match_count, and filter. You must create this match function first — see the Supabase pgvector docs for the standard match_documents example.


Setting up the block

  1. Add the Supabase block (Database category) to your workflow.
  2. Select or create your Supabase Project credentials (Project URL + API Key).
  3. Pick an Action from the dropdown.
  4. Fill in the fields shown for that action. Every field accepts {{variables}}.
  5. Connect the block's output to the next step.

Example: Save a lead to a table

A form webhook triggers the workflow. Add a Supabase block:

  • Action: Insert Row
  • Table: leads
  • Row Data (JSON): {"name": "{{name}}", "email": "{{email}}", "source": "website"}

Every submission becomes a new row in your leads table. Add a second Supabase block with Auth: Invite User to send the new lead a sign-up invite.


Output

The block returns the Supabase API response. What you get depends on the operation:

Operation typeOutput
Row operationsThe affected row(s) as JSON
Auth operationsThe created user object; Generate Magic Link also returns action_link
Storage: Get Signed URLThe temporary signed URL
Storage: List ObjectsThe list of objects
RPC / Edge FunctionWhatever your function returns
Vector SearchThe matched rows with similarity scores
Indite Documentation v1.7.1
PrivacyTermsSupport