Supabase MCP Server on Phala Cloud

What is Supabase MCP Server on Phala Cloud?

This server hosts the Supabase MCP server in a Trusted Execution Environment (TEE) on Phala Cloud. It is a fork of the official supabase-community MCP server, adding Docker support and SSE transport for remote hosting and easy deployment.

How to use Supabase MCP Server on Phala Cloud?

Deploy on Phala Cloud by copying the provided docker-compose.yml to the Phala Cloud deployment dashboard and setting environment variables. Alternatively, build locally using docker build --platform linux/amd64 -t <image> . and run with docker compose up after setting SUPABASE_ACCESS_TOKEN. Verify the server using npx @modelcontextprotocol/inspector and connect via SSE URL http://localhost:3000/sse.

Key features of Supabase MCP Server on Phala Cloud

• Hosts Supabase MCP in TEE on Phala Cloud
• Adds Docker support for easy deployment
• Supports SSE transport for remote access
• Forked from official supabase-community MCP server

Use cases of Supabase MCP Server on Phala Cloud

• Deploy a secure Supabase MCP server in a TEE environment
• Remotely access Supabase MCP capabilities via SSE
• Run the server locally with Docker for development

FAQ from Supabase MCP Server on Phala Cloud

What is the difference from the official Supabase MCP server?

This fork adds Docker support and SSE transport, enabling deployment on Phala Cloud and remote hosting via HTTPS/SSE.

What runtime dependencies are required?

Docker and a Phala Cloud account (for cloud deployment) or a local Docker environment for self-hosting.

How do I authenticate with the server?

Set the environment variable SUPABASE_ACCESS_TOKEN with your Supabase access token before launching the server.

What transport protocol does the server use?

The server uses SSE (Server-Sent Events) transport, accessible at http://localhost:3000/sse when running locally.

Can I deploy this on my own infrastructure?

Yes, you can build and run the Docker image locally using the provided Dockerfile and docker-compose.yml, or deploy to any Docker-compatible platform.