How to integrate your Agent with Supabase

​

Overview

​

Advantages of Using Supabase with Release0 Agent

• Store and retrieve user data in a robust PostgreSQL database
• Create responsive flows with real-time database-driven logic
• Adapt dialogs based on stored user preferences and history
• Create new records (e.g., user registrations, form submissions)
• Read data for personalization and decision-making
• Update existing records based on conversation outcomes
• Delete entries when necessary

Example: Store user preferences and retrieve them in future conversations using {{user.preference_id}}.

​

Setting Up Supabase Authentication

​

1. Supabase Project URL

https://your-project-id.supabase.co

​

2. Service Role API Key

1. Go to your Supabase dashboard
2. Navigate to Settings > API Keys
3. Copy the service_role key under “Project API keys”

​

Available Actions

​

1. Create Row

• Table Name: The name of the table to insert into
• Fields: Key-value pairs for the data to insert

• Schema: Database schema (defaults to “public”)
• Response Mapping: Map returned data to variables

​

2. Get Rows

• Table Name: The name of the table to query

• Schema: Database schema (defaults to “public”)
• Return Type: Choose “All”, “First”, “Last”, or “Random”
• Order By Column: Column to sort results
• Order Direction: “asc” (ascending) or “desc” (descending)
• Filter Conditions: Add WHERE conditions
• Select Columns: Specify which columns to return (defaults to ”*”)
• Limit: Maximum number of records to return
• Response Mapping: Map returned data to variables

​

3. Update Rows

• Table Name: The name of the table to update
• Filter Conditions: Conditions to identify which records to update
• Fields to Update: Key-value pairs for the data to update

• Schema: Database schema (defaults to “public”)
• Response Mapping: Map updated data to variables

​

4. Delete Rows

• Table Name: The name of the table to delete from
• Filter Conditions: Conditions to identify which records to delete

• Schema: Database schema (defaults to “public”)
• Response Mapping: Map deleted data to variables

​

Complete Workflow Example

​

Step 1: Create a Support Tickets Table

• id: Primary key (auto-increment)
• customer_name: Text
• customer_email: Text
• issue_type: Text
• description: Text
• status: Text (default: “open”)
• created_at: Timestamp (default: now())
• priority: Text (default: “medium”)

​

Step 2: Collect Customer Information

Hi! I'm here to help you create a support ticket. 

Please provide the following information:
- Your name: {{customer_name}}
- Your email: {{customer_email}}  
- Issue type: {{issue_type}}
- Description of the problem: {{description}}
- Priority level (low/medium/high): {{priority}}

​

Step 3: Create the Support Ticket

• Table Name: support_tickets
• Fields:
  • customer_name: {{customer_name}}
  • customer_email: {{customer_email}}
  • issue_type: {{issue_type}}
  • description: {{description}}
  • priority: {{priority}}
  • status: open

• Map “Row Created” to variable {{ticket_data}}
• Map “Table Name” to variable {{table_name}}

​

Step 4: Confirmation Message

✅ Your support ticket has been created successfully!

Ticket Details:
- Ticket ID: {{ticket_data.id}}
- Status: {{ticket_data.status}}
- Priority: {{ticket_data.priority}}
- Created: {{ticket_data.created_at}}

We'll get back to you at {{customer_email}} soon!

​

Step 5: Ticket Status Lookup

Please enter your ticket ID to check the status: {{ticket_id}}

• Table Name: support_tickets
• Filter Conditions: id equals {{ticket_id}}
• Return Type: “First”

• Map “Rows Found” to variable {{ticket_info}}

📋 Ticket Status Report

Ticket ID: {{ticket_info.id}}
Customer: {{ticket_info.customer_name}}
Issue Type: {{ticket_info.issue_type}}
Status: {{ticket_info.status}}
Priority: {{ticket_info.priority}}
Created: {{ticket_info.created_at}}

Description: {{ticket_info.description}}

​

Step 6: Update Ticket Status (Admin Flow)

• Table Name: support_tickets
• Filter Conditions: id equals {{ticket_id}}
• Fields to Update:
  • status: {{new_status}}

​

Best Practices

​

Security

• Always use the service_role key for server-side operations
• Implement Row Level Security (RLS) policies in Supabase
• Validate user inputs before database operations

​

Performance

• Use specific column selection instead of SELECT * when possible
• Add appropriate indexes for frequently filtered columns
• Use pagination with LIMIT for large datasets

​

Error Handling

• Always include filter conditions for UPDATE and DELETE operations
• Test your database operations thoroughly before deployment
• Use Response Mapping to capture and handle operation results

​

Data Validation

• Validate required fields before database operations
• Use appropriate data types in your Supabase schema
• Implement constraints and default values in your database

​

Filter Conditions

• Equals: Exact match
• Not Equals: Exclude exact matches
• Greater Than: Numeric/date comparisons
• Less Than: Numeric/date comparisons
• Contains: Text search within fields
• Starts With: Text prefix matching
• Ends With: Text suffix matching
• Is Empty: Check for null/empty values
• Is Not Empty: Check for non-null values

​

Response Mapping Options

• Row Created: The newly created record
• Table Name: Name of the table

• Rows Found: Array of matching records
• Table Name: Name of the table
• Rows Count: Number of records found

• Updated Rows: Array of updated records
• Table Name: Name of the table
• Updated Count: Number of records updated

• Deleted Rows: Array of deleted records
• Table Name: Name of the table
• Deleted Count: Number of records deleted

​

Troubleshooting

​

Common Issues

1. Authentication Errors: Ensure you’re using the service_role key, not the anon key
2. Table Not Found: Verify table names are correct and case-sensitive
3. Permission Denied: Check your RLS policies if enabled
4. Row Level Security (RLS) Policy Violations: If you get an error like "new row violates row-level security policy for table" with a 401 Unauthorized status, you need to either:
   • Disable RLS for your table (go to Authentication > Policies in Supabase dashboard)
   • Create appropriate RLS policies that allow your service_role key to perform operations
   • For testing purposes, you can temporarily disable RLS by running: ALTER TABLE your_table_name DISABLE ROW LEVEL SECURITY;
5. Filter Errors: Ensure filter conditions use correct column names and data types

​

Debugging Tips

• Use the browser’s developer tools to inspect network requests
• Check Supabase logs in your dashboard for detailed error messages
• Test database operations directly in Supabase SQL editor first
• Verify your table schema matches the data you’re sending

​

Validation and Testing

1. Test Each Operation: Verify Create, Read, Update, and Delete operations work correctly
2. Validate Filters: Ensure filter conditions return expected results
3. Check Response Mapping: Confirm mapped variables contain the correct data
4. Test Error Scenarios: Verify graceful handling of invalid inputs or database errors
5. Performance Testing: Test with realistic data volumes

💡 Want to see this in action? Check out our How to Build Data-Driven Chatbots with Supabase Integration blog post for a real-world example and practical tips.