Supabase Automation

What it does

This skill enables AI agents to automate various Supabase operations through Composio's Supabase toolkit, leveraging Rube MCP for tool integration. It allows for database queries, table schema inspection, SQL execution, project and organization management, storage bucket manipulation, edge function handling, and service health monitoring. The agent can interact with these functionalities by requesting specific tools and parameters via the Rube MCP interface.

When to use it

• When you need to retrieve data from a Supabase database table.
• To inspect the schema of a Supabase table (column types, constraints).
• For executing complex SQL queries beyond simple selections or reads within your Supabase project.
• To manage and monitor different aspects of your Supabase project, such as storage buckets or edge functions.

Key capabilities

• Listing all projects in your Supabase account (SUPABASE_LIST_ALL_PROJECTS).
• Listing tables and views within a database (SUPABASE_LIST_TABLES).
• Retrieving detailed table schemas (SUPABASE_GET_TABLE_SCHEMAS).
• Querying data from tables with filtering, sorting, and pagination (SUPABASE_SELECT_FROM_TABLE).
• Executing arbitrary SQL queries for inserts, updates, or deletes (SUPABASE_BETA_RUN_SQL_QUERY).

Example prompts

• "List all projects in my Supabase account."
• "What is the schema of the 'users' table in my database?"
• "Run a query to select all users with email addresses ending in '@example.com'."
• "Execute this SQL statement: INSERT INTO products (name, price) VALUES ('New Product', 29.99);"

Tips & gotchas

• Rube MCP is required: Ensure Rube MCP (https://rube.app/mcp) is added as an MCP server in your client configuration and that RUBE_SEARCH_TOOLS responds before using this skill.
• Connection Required: You must establish an active Supabase connection via RUBE_MANAGE_CONNECTIONS with the 'supabase' toolkit. Follow any authentication links provided if the connection is not already active.
• Tool Order Matters: Always call RUBE_SEARCH_TOOLS first to get current tool schemas, and follow the specified tool sequences for optimal results (e.g., listing projects before querying tables).