Mem0 (Long Term Memory)

🚀 MCP-Mem0: A Long-Term Memory System for AI Agents

MCP-Mem0 is a long-term memory system built on the Model Context Protocol (MCP). It provides a framework for AI agents to store and retrieve long-term information, helping AI maintain consistency and coherence across tasks.

🚀 Quick Start

MCP-Mem0 offers a robust solution for AI agents to manage long - term memory. By following the steps in this README, you can quickly set up and start using this system.

✨ Features

1. Support for Multiple Tool Integrations: Easily integrate different AI tools and services through the MCP protocol.
2. Persistent Storage: Use a database to store memory content, ensuring information is not lost.
3. Scalable Architecture: Facilitate the addition of custom tools and functional modules.
4. Intuitive API Interface: Provide a simple and easy - to - use programming interface for developers.

📦 Installation

Prerequisites

• Python 3.8 or higher
• Basic development environment (e.g., pip, venv, etc.)
• Install PostgreSQL or other supported database systems

Installation Methods

Install using pip

pip install mcp-mem0

Clone the repository using Git and install

git clone https://github.com/yourusername/mcp-mem0.git
cd mcp-mem0
python setup.py install

📚 Documentation

Configuration Guide

Create a configuration file

Create a config.json file in the root directory of the project:

{
  "database": {
    "url": "postgresql://localhost:5432/mem0",
    "username": "mem0_user",
    "password": "your_password"
  },
  "logging": {
    "level": "INFO",
    "filename": "mcp-mem0.log"
  }
}

Set environment variables

export DATABASE_URL=postgresql://localhost:5432/mem0
export LOG_LEVEL=DEBUG

Running Guide

Start in development mode

python -m mcp_mem0.cli --mode development

Run in production mode

gunicorn --bind :8000 mcp_mem0.app:app

Custom Tool Development

Create a new tool module

Create a new Python file in the tools/ directory, for example:

from mcp_core import tool

@tool(name="custom_tool", description="Execute custom tasks")
def custom_function(context):
    # Your implementation code
    pass

Register the tool with the system

Import and register the tool in the main application:

from mcp_mem0.core import MCPSystem
from .tools import custom_function

mcp = MCPSystem()
mcp.register_tool(custom_function)

💻 Usage Examples

Basic Usage

from mcp_mem0.client import MCPClient

client = MCPClient('localhost', 8000)
result = client.invoke_tool('custom_tool', parameters)
print(result)

Advanced Usage

from mcp_mem0.prompt import create_prompt
from mcp_mem0.resource import load_resource

prompt = create_prompt("Answer the user's question: {}", ["Common sense", "Humor"])
resource = load_resource("data/custom_dataset.csv")
result = client.invoke_with_context(prompt, resource)

Troubleshooting

Common Errors and Solutions

1. Failed to connect to the database
   • Check if the database is running.
   • Ensure the credential information is correct.
2. Slow API response
   • Adjust query parameters to optimize data retrieval logic.
3. Tool registration issues
   • Make sure the tool module is correctly imported.
   • Check the log output for more information.

More Resources

• Official Documentation
• GitHub Repository
• Community Discussion Group

By following the above steps, you can easily set up and expand your own MCP - Mem0 long - term memory system. If you have any questions or suggestions, please feel free to contact the project maintenance team.