Looking for a complete MCP server installation guide? This tutorial walks you through every step — from checking prerequisites to configuring AI Memory's MCP server with Claude Desktop, Cursor, VS Code, and Windsurf. By the end, your AI assistants will be able to search, add, and retrieve context from all your saved conversations through the Model Context Protocol.

Prerequisites

Before you begin this MCP server installation guide, make sure you have the following ready:

Check Your Python Version

Open your terminal and run the following command to verify your Python version:

python --version

You should see Python 3.10.x or higher. If you have Python 3.9 or lower, upgrade first. On macOS with Homebrew: brew install python@3.12. On Ubuntu/Debian: sudo apt install python3.12. On Windows, download the latest installer from python.org.

Step 1 — Install aimemory-mcp-server

With Python ready, install the AI Memory MCP server package from PyPI with a single command:

pip install aimemory-mcp-server

This downloads and installs the aimemory-mcp-server package along with all its dependencies. The package provides:

• An MCP server exposing tools: search_memory, add_memory, get_context, and list_memories
• SQLite-based local storage for conversations
• Stdio transport (default) and HTTP/SSE transport for remote access
• Cross-platform support (macOS, Windows, Linux)

Verify the Installation

After installation, verify the server is available:

aimemory-mcp-server --help

You should see usage information printed to the terminal. If you get a "command not found" error, ensure your Python scripts directory is in your system PATH. On most systems this happens automatically.

Optional: Install in a Virtual Environment

To keep your system Python clean, you can create a virtual environment first:

python -m venv ~/.aimemory-venv
source ~/.aimemory-venv/bin/activate  # macOS/Linux
# or: ~/.aimemory-venv/Scripts/activate  # Windows
pip install aimemory-mcp-server

If you use a venv, note the full path to the aimemory-mcp-serverbinary inside the venv's bin/directory — you'll need it when configuring your MCP client.

Step 2 — Configure Claude Desktop

Claude Desktop has native MCP support. To add the AI Memory MCP server:

1. Locate your Claude Desktop configuration file:
   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json
2. Open the file in a text editor and add the MCP server configuration:

{
  "mcpServers": {
    "ai-memory": {
      "command": "aimemory-mcp-server",
      "args": []
    }
  }
}

3. Save the file and restart Claude Desktop completely.
4. You should see the AI Memory MCP server listed in Claude's MCP servers section.

Step 3 — Configure Cursor

Cursor supports MCP servers through its settings. Here's how to set it up:

1. Open Cursor and go to Settings → Cursor Settings → MCP
2. Click "Add new global MCP server"
3. Paste the following configuration:

{
  "mcpServers": {
    "ai-memory": {
      "command": "aimemory-mcp-server",
      "args": []
    }
  }
}

Alternatively, you can edit the MCP configuration file directly at ~/.cursor/mcp.json (global) or .cursor/mcp.json (per-project). After saving, Cursor will automatically detect and connect to the MCP server.

Step 4 — Configure VS Code (with Continue or Copilot)

VS Code supports MCP through extensions like Continueand GitHub Copilot. To configure AI Memory's MCP server in VS Code:

1. Install the Continue extension from the VS Code marketplace (or use Copilot Chat if you have MCP support enabled)
2. Open your MCP configuration. For Continue, edit ~/.continue/config.json or use the Continue settings UI
3. Add the MCP server entry:

{
  "mcpServers": [
    {
      "name": "ai-memory",
      "command": "aimemory-mcp-server",
      "args": []
    }
  ]
}

For GitHub Copilot, add the MCP server to your workspace .vscode/settings.json:

{
  "github.copilot.chat.mcp.servers": {
    "ai-memory": {
      "command": "aimemory-mcp-server",
      "args": []
    }
  }
}

Restart VS Code after saving. The AI Memory tools will be available in your AI chat sessions.

Step 5 — Configure Windsurf

Windsurf (by Codeium) supports MCP servers natively. To add AI Memory:

1. Open Windsurf and go to Settings → Cascade → MCP Servers
2. Click "Add Server"
3. Enter the following configuration:

{
  "mcpServers": {
    "ai-memory": {
      "command": "aimemory-mcp-server",
      "args": []
    }
  }
}

You can also manually edit the Windsurf MCP configuration file at ~/.windsurf/mcp.json. After saving, restart Windsurf and the server will be available in Cascade sessions.

HTTP/SSE Transport for Remote Access

By default, the MCP server uses stdio transport — it runs as a local subprocess of your AI client. This is the simplest and most secure option for individual use. However, if you need to share the MCP server across a team, run it on a remote machine, or access it from multiple devices, you can switch to HTTP/SSE transport.

Starting the Server in HTTP Mode

Set the AIMEMORY_TRANSPORT environment variable to http and start the server:

AIMEMORY_TRANSPORT=http AIMEMORY_PORT=8080 aimemory-mcp-server

The server will start listening on http://localhost:8080. You can change the port with the AIMEMORY_PORT variable.

Connecting Remote Clients

Remote MCP clients can connect via the HTTP Streamable or SSE endpoint. In your client configuration, use the url field instead of command:

{
  "mcpServers": {
    "ai-memory-remote": {
      "url": "http://your-server:8080/mcp"
    }
  }
}

Environment Variables Reference

The AI Memory MCP server is configured through environment variables. Here is the complete reference:

Examples

Custom database location:

AIMEMORY_DB=/data/aimemory/mydb.db aimemory-mcp-server

Run as HTTP server on a custom port:

AIMEMORY_TRANSPORT=http AIMEMORY_PORT=9090 aimemory-mcp-server

You can also set these variables in a .env file or export them in your shell profile (~/.bashrc, ~/.zshrc, etc.) for persistent configuration.

Troubleshooting Common Issues

Encountering problems? Here are the most common issues and their solutions when following this MCP server installation guide:

Frequently Asked Questions

Next Steps

Congratulations — your MCP server installation is complete! Here's what to do next:

• Import existing conversations — Use the AI Memory browser extension to export your ChatGPT, Claude, and DeepSeek conversations, then import them into the MCP server database
• Try searching — Ask your AI assistant to "search my memory" for a topic you've discussed before
• Set up remote access — Switch to HTTP transport if you want to share the server across devices or team members
• Explore more MCP servers — Browse the growing ecosystem of MCP servers for databases, APIs, file systems, and more