Membase MCP

🚀 Unibase membase Server

Unibase membase is the first decentralized memory layer designed for AI agents. It stores historical information, interaction records, and persistent data of agents, ensuring their continuity and traceability.

🚀 Quick Start

Unibase membase is the first decentralized memory layer crafted for AI agents. It stores historical data, interaction logs, and persistent agent data, ensuring seamless continuity and traceability.

The Membase mcp server seamlessly integrates with the membase protocol for decentralized storage. It enables uploading and downloading of current conversation memories to the Unibase network.

✨ Features

Messages or memories can be accessed via the following address: https://testnet.hub.membase.io/

• get_conversation_id: Retrieve the current conversation ID.
• switch_conversation: Switch to a different conversation.
• save_message: Save a message/memory to the current conversation.
• get_messages: Fetch the last n messages from the current conversation.

📦 Installation

git clone https://github.com/unibaseio/membase-mcp.git
cd membase-mcp
uv run src/membase_mcp/server.py

📚 Documentation

Environment Variables

• MEMBASE_ACCOUNT: Your account used for uploading.
• MEMBASE_CONVERSATION_ID: The current conversation ID, which should be unique and its history will be pre - loaded.
• MEMBASE_ID: Your instance ID.

Configuration (for Claude/Windsurf/Cursor/Cline)

{
  "mcpServers": {
    "membase": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/membase-mcp",
        "run", 
        "src/membase_mcp/server.py"
      ],
      "env": {
        "MEMBASE_ACCOUNT": "Your account, 0x...",
        "MEMBASE_CONVERSATION_ID": "The current conversation ID, should be unique and its history will be pre - loaded",
        "MEMBASE_ID": "Your sub - account, any string"
      }
    }
  }
}

💻 Usage Examples

Basic Usage

Invoke functions during LLM chats:

• Get the conversation ID and switch conversations

• Save messages and retrieve messages