Memory Cloud Protocol (MCP) Server

What is Memory Cloud Protocol (MCP) Server?

Memory Cloud Protocol (MCP) Server is a backend service for persisting and coordinating agent memory, designed for use with AI agents such as Cursor. It stores memory records, API tokens, project states, and function deployment metadata in a Supabase/PostgreSQL database, providing automatic entity lookup and REST endpoints for agent coordination.

How to use Memory Cloud Protocol (MCP) Server?

Clone the repository, install dependencies with npm install, copy .env.example to .env, set up a Supabase project with the required tables, then start the server using npm run dev (development) or npm start (production). Use the provided test scripts (test-local.sh, test-railway.sh) to verify health endpoints and authentication.

Key features of Memory Cloud Protocol (MCP) Server

• Persists memory records for Supabase projects including UUIDs, table names, RLS states, and keys
• Stores API tokens, project states, and function deployment metadata
• Logs agent tasks and output
• Provides automatic lookup of known entities to prevent repetitive errors
• Exposes REST endpoints for memory management and agent coordination

Use cases of Memory Cloud Protocol (MCP) Server

• Maintaining persistent context for AI agents across sessions
• Coordinating agent actions through dispatch instructions
• Logging and reviewing agent errors and task outputs

FAQ from Memory Cloud Protocol (MCP) Server

What database does Memory Cloud Protocol (MCP) Server use?

The server uses Supabase/PostgreSQL. You must create three tables (memory, logs, and dispatches) along with a UUID extension in your Supabase project.

What runtime dependencies are required?

Node.js is required. After cloning, run npm install to install dependencies. A Supabase project with the required tables must be available.

Where does data persist?

All memory records, logs, and dispatches are stored in the Supabase PostgreSQL database you configure.

How is the server deployed?

It is designed for Railway.app deployment using a Dockerfile and railway.json configuration. The server can also be run locally with Docker (docker build -t mcp-server then docker run -p 3002:3002 --env-file .env mcp-server).

What authentication is used?

Most REST endpoints require an MCP_API_KEY for authentication. The /health endpoint is public and does not require authentication.