supabase-automation

Supabase Automation - Claude Database Automation Management Skills

Skill Overview

Using Composio’s Rube MCP in Claude to automatically manage Supabase databases, table structures, project configurations, storage buckets, and edge functions—so you can complete querying, monitoring, and operational tasks without switching tools.

Use Cases

1. Database Querying and Maintenance

When you need to quickly query Supabase table data, inspect table schemas, perform SQL updates, or run batch operations, you can do it directly in a Claude conversation—no need to log into the Supabase Dashboard or write scripts.

2. Project Operations and Monitoring

When you need to check the health status of Supabase project services, review API configurations, monitor database branches, or manage organization members, this skill lets you get all critical information with one click.

3. Development Workflow Integration

During development, when you need to generate TypeScript types, check the list of storage buckets, or review the status of Edge Functions, integrate this skill into your workflow to improve development efficiency.

Core Features

1. Database Querying and Operations

Use PostgREST filters (eq, gt, like, cs, etc.) for precise data queries, supporting pagination, sorting, and nested field selection. For complex operations, you can directly execute SQL statements, including DML operations such as INSERT, UPDATE, and DELETE.

2. Table Structure and Schema Management

List all tables and views, retrieve detailed column types, constraints, and relationships, and automatically generate TypeScript type definitions. Ideal for understanding database structure before development or for generating types for frontend code.

3. Project and Organization Management

List all organizations and projects, retrieve database and authentication configurations, check service health status (auth, db, storage, etc.), view Edge Functions and storage bucket lists—manage Supabase resources end-to-end.

Common Questions

How do I connect Supabase in Claude?

First, configure the Rube MCP server (add https://rube.app/mcp to your MCP configuration). Then, in the conversation, call RUBE_MANAGE_CONNECTIONS to select the supabase tool pack. If the connection isn’t active, it will return an authentication link. After completing authentication, you can use all Supabase features.

How do I get the Supabase project’s project_ref?

Use SUPABASE_LIST_ALL_PROJECTS to list all projects. Each project includes a ref field, which is a 20-character lowercase letter project reference (format like abcdefghijklmnopqrst). This ref is required when querying tables, running SQL, and other operations.

What operations does the Supabase MCP skill support?

Supports database querying (SELECT), SQL execution, table schema queries, viewing project configuration, managing organization members, checking service health, listing Edge Functions, listing storage buckets, generating TypeScript types, and more. Most operations are read-only; write operations require execution via SQL.

How do I query data using PostgREST filters?

When calling SUPABASE_SELECT_FROM_TABLE, specify column, operator, and value in the filters parameter. Common operators include eq (equals), neq (not equals), gt/gte (greater than/greater than or equal), like/ilike (fuzzy match), in (included in a list), cs (array contains), etc.

Can Supabase automation execute SQL write operations?

Yes. Use SUPABASE_BETA_RUN_SQL_QUERY to execute any SQL statement, including DML operations such as INSERT, UPDATE, and DELETE. It’s recommended to set read_only: true for SELECT queries to improve safety. Note that SQL execution has an ~60-second timeout limit.

What is the project_ref format?

project_ref is the unique identifier of a Supabase project, consisting of exactly 20 lowercase English letters (a-z). It contains no numbers or hyphens, formatted like ^[a-z]{20}$. An invalid format will cause API calls to fail.