AI Memory That Can Be Shared With Your Team

A Model Context Protocol (MCP) server that gives your AI assistants a permanent memory they can share across all your tools and with your entire team.

Get Started View Source

Key Features

📂 Project-Specific Memory

Automatically detects project contexts and stores memories within your project's directory structure.

🧠 Structured Storage

Organizes knowledge into Entities, Concepts, and Sessions for better retrieval and context.

🔍 Full-Text Search

🤝 Multi-Tool Compatible

Works seamlessly with Claude Desktop, Cursor, Windsurf, Cline, and any other MCP client.

🔒 Git Integration

Memories are stored as Markdown files, making them easy to commit, review, and share with your team.

⚙️ Flexible Config

Configure directory paths and settings via simple JSON configuration files.

Quick Start

Install the server and configure it with your favorite MCP tool.

1. Installation

# Install dependencies
npm install

# Build the project
npm run build

2. Configuration

Add this to your tool's mcpServers configuration:

{
  "mcpServers": {
    "shared-memory": {
      "command": "npx",
      "args": ["-y", "mcp-shared-memory"]
    }
  }
}

Configuration

Create a configuration file in your project root to customize behavior.

Recommended (.mcp-memory.json)

{
  "memoryDir": ".context/memory",
  "enabled": true,
  "description": "My project"
}

Visible (mcp-memory.json)

{
  "memoryDir": "memory", 
  "enabled": true,
  "description": "My project"
}

Documentation & Resources

Document	Description
USAGE.md	Complete usage guide with setup for all MCP tools
CHANGELOG.md	Version history and release notes
instructions_template.md	AI assistant instructions template
CONTRIBUTING.md	Contribution guidelines and workflow

Contributing

We welcome contributions to help improve this project! Whether it's fixing bugs, adding new features, or improving documentation.

Please review the CONTRIBUTING.md file for detailed guidelines on how to contribute.

Open Pull Request