Memberberry LogoMemberberry
Back to home

Documentation

Connect Memberberry to all your AI tools in minutes using the Model Context Protocol

Quick Start

Get Your API Key

You'll need this to authenticate with Memberberry

  1. 1Sign up for a free account
  2. 2Copy your API key from the dashboard
  3. 3Add the MCP configuration to your client

Setup Guides

Claude Desktop

Add shared memory to Claude Desktop conversations

1

Open Connectors Settings

In Claude Desktop:

  1. Click the menu button (☰) in the bottom left
  2. Select "Connectors" or go to Settings → Connectors
  3. Click "Add custom connector" at the bottom
Claude Desktop Connectors Settings
2

Configure Memberberry Connection

Fill in the connector details:

Name:

Memberberry

Remote MCP server URL:

https://memberberry.radicalgeek.co.uk/mcp
Add Memberberry connector in Claude Desktop

💡Claude Desktop uses OAuth for secure authentication. You can either use Dynamic Client Registration (automatic) or manually create an OAuth app.

3

Authenticate with OAuth

Click "Add" and Claude Desktop will:

  1. Automatically register as an OAuth client (Dynamic Client Registration)
  2. Open your browser to authorize the connection
  3. Log in to your Memberberry account if prompted
  4. Click "Approve" to grant access
  5. Return to Claude Desktop - connection complete!

Optional: You can also manually create an OAuth app in the dashboard and enter the Client ID and Client Secret in Advanced settings.

OAuth configuration in Claude Desktop

OAuth provides automatic token refresh and more granular permissions.

4

Verify Connection

You should see Memberberry listed in your active connectors.

Memberberry successfully connected in Claude Desktop

Start a conversation and Memberberry will automatically store and recall important information!

Cursor

Enable persistent memory in your AI-powered code editor

1

Open Cursor Settings

Go to Settings → Features → Model Context Protocol (or press ⌘, on Mac)

2

Add Memberberry Server

Add the Memberberry MCP configuration:

{
  "mcpServers": {
    "memberberry": {
      "url": "https://memberberry.radicalgeek.co.uk/mcp",
      "headers": {
        "x-api-key": "your-api-key-here",
        "Accept": "application/json, text/event-stream"
      },
      "type": "http"
    }
  }
}

Replace with your API key from the dashboard.

3

Start Coding with Memory

Cursor will now remember project decisions, architecture patterns, and your coding preferences across all conversations.

VS Code

Add memory to GitHub Copilot and other AI assistants

1

Configure Memberberry

Add to your MCP configuration:

{
  "mcpServers": {
    "memberberry": {
      "url": "https://memberberry.radicalgeek.co.uk/mcp",
      "type": "http"
    }
  }
}

That's it! OAuth authentication will be handled automatically when you first connect.

Any MCP-Compatible Tool

Universal setup for all MCP clients

1

MCP Configuration

All MCP clients can connect to Memberberry using this standard configuration:

Endpoint URL:

https://memberberry.radicalgeek.co.uk/mcp

Authentication Header:

x-api-key: your-api-key-here

Full Configuration Example:

{
  "mcpServers": {
    "memberberry": {
      "url": "https://memberberry.radicalgeek.co.uk/mcp",
      "headers": {
        "x-api-key": "your-api-key-here",
        "Accept": "application/json, text/event-stream"
      },
      "type": "http"
    }
  }
}

How It Works

Using Memories in Chat

Memberberry works seamlessly through natural conversation

Storing Memories

Simply chat naturally! When you share important information, ask the AI to remember it:

You:

"I prefer using TypeScript for all new projects. Please remember this."

AI:

✓ I've stored that preference in your memories.

Retrieving Memories

The AI automatically recalls relevant memories based on your conversation context:

You:

"Let's start a new project"

AI:

Based on your preferences, I'll set up a TypeScript project...

Searching Memories

Ask the AI to search your memories anytime:

You:

"What are my coding preferences?"

"Search my project memories"

"What do I remember about the SHAPO project?"

VS Code: Configure Tool Sets (Optional)

Make Memberberry tools easily accessible in GitHub Copilot Chat

Press ⌘⇧P (Mac) or Ctrl+Shift+P (Windows/Linux) and run:

> Chat: Configure Tool Sets...

Then add a toolset.jsonc file with:

{
  "tools": [
    {
      "name": "store_memory",
      "description": "Store important information for long-term memory",
      "parameters": {
        "content": "The information to remember",
        "category": "Type: facts, preferences, conversations, projects, learning, goals, context, or reminders"
      }
    },
    {
      "name": "search_memory",
      "description": "Search stored memories using text or semantic search",
      "parameters": {
        "query": "What to search for",
        "category": "(optional) Filter by category"
      }
    },
    {
      "name": "list_memories",
      "description": "List recent memories, optionally filtered by category"
    },
    {
      "name": "get_stats",
      "description": "View memory usage statistics"
    }
  ]
}

This makes Memberberry tools appear in your Copilot Chat suggestions!

Categories

  • facts: Factual information
  • preferences: User preferences and settings
  • conversations: Important context from conversations
  • projects: Project-specific information
  • learning: Things to remember for later

Team Collaboration

Share memories with your team. Everyone on your account can access and contribute to the shared memory pool.