This project provides a server that connects AI assistants to Supabase projects, enabling them to manage data, configurations, and project settings through the Model Context Protocol (MCP). It targets developers and AI users who want to integrate their Supabase backend with LLM-powered workflows, offering a standardized way for AI to interact with their database and project infrastructure.

How It Works

The server acts as a bridge, translating AI assistant commands into Supabase API calls. It leverages Node.js and npx to download and run the latest version of the Supabase MCP server. The core mechanism involves a Supabase Personal Access Token (PAT) for authentication, allowing the AI to securely access and manipulate project resources. Configuration is managed via JSON within the AI client, specifying the command to execute the server.

Quick Start & Requirements

• Install Node.js (version check: node -v).
• Obtain a Supabase Personal Access Token (PAT).
• Configure your MCP client (e.g., Cursor) with the provided JSON structure, replacing <personal-access-token> with your PAT.
• For project-specific access, use the --project-ref flag. For read-only access, use the --read-only flag.
• Official Docs: Model Context Protocol

Highlighted Details

• Supports a wide range of Supabase tools, including project management, database operations, Edge Function management, and branching (experimental).
• Offers project-scoped and read-only modes for granular access control.
• Includes tools for generating TypeScript types from database schemas.
• Provides cost estimation and confirmation for project/branch creation.

Maintenance & Community

• The project is marked as pre-1.0, indicating potential breaking changes.
• License: Apache 2.0.

Licensing & Compatibility

• Apache 2.0 license permits commercial use and integration with closed-source projects.

Limitations & Caveats

• The project is pre-1.0, so breaking changes are expected.
• Branching features are experimental and may require a paid Supabase plan.
• Account-level tools are unavailable when the server is scoped to a specific project.