Mem0 MCP

🚀 Mem0 MCP Server Configuration Guide

This guide provides a detailed introduction to the Mem0 MCP server, including installation, configuration, usage examples, and advanced features.

🚀 Quick Start

To run the Mem0 MCP server, follow these steps:

# Install dependencies
npm install mem0ai safeLogger

# Start the server
node src/index.js --mode [cloud/local] [--userId <default user ID>]

✨ Features

• Dual Storage Modes: Supports both cloud and local storage modes to meet different needs.
• Core Tools: Provides add_memory and search_memory tools for easy memory management.
• Advanced Parameters: Allows for fine - tuning of memory addition and search operations.
• Comprehensive Error Handling: Offers solutions for common errors.
• Performance Optimization: Provides methods for optimizing server performance.

📦 Installation

To install the Mem0 MCP server, you need to install the necessary dependencies using the following command:

npm install mem0ai safeLogger

💻 Usage Examples

Adding Memory

{
  "content": "Important information to remember",
  "userId": "user123",
  "sessionId": "project-abc",
  "metadata": {
    "custom_instructions": "Process this memory with priority"
  }
}

Searching Memory

{
  "query": "Recent project updates",
  "userId": "user123",
  "filters": {
    "AND": [
      { "field": "project_id", "value": "abc123" },
      { "field": "timestamp", "operator": ">", "value": "2023-01-01" }
    ]
  },
  "top_k": 5
}

📚 Documentation

Configuration Options

The MCP server supports two operating modes:

• Cloud Storage Mode: Uses Mem0's cloud services to store and retrieve memories. You need to set the MEM0_API_KEY environment variable.
• Local Storage Mode: Processes memories locally using OpenAI embeddings. You need to set the OPENAI_API_KEY environment variable.

Core Tools

add_memory

• Description: Adds new memory content to the storage.
• Input Format:

{
  "content": string,
  "userId": string,
  "sessionId": string (optional),
  "metadata": object (optional)
}

• Output: Returns a success status or an error message.

search_memory

• Description: Retrieves relevant memories from the storage based on the query.
• Input Format:

{
  "query": string,
  "userId": string,
  "filters": object (optional),
  "top_k": number (optional, default 10),
  "threshold": number (optional, default 0.3)
}

• Output: Returns a list of matching memories and similarity scores.

Advanced Parameters

Advanced Options for add_memory

{
  "content": string,
  "userId": string,
  "sessionId": string (optional),
  "metadata": {
    "includes": string, // Specific inclusion preferences
    "excludes": string, // Specific exclusion preferences
    "immutable": boolean, // Whether it is immutable
    "expiration_date": string, // Expiration date, format: YYYY - MM - DD
    "custom_instructions": string, // Project - specific instructions
    "version": string // Memory version, default is v2
  }
}

Advanced Options for search_memory

{
  "query": string,
  "userId": string,
  "filters": {
    "OR": [ array of conditions ] | 
    "AND": [ array of conditions ]
  },
  "top_k": number, // Maximum number of returned results
  "threshold": number, // Similarity threshold
  "fields": string[], // Specific fields to return
  "rerank": boolean, // Whether to re - rank, default is false
  "keyword_search": boolean, // Whether to perform keyword search, default is false
  "filter_memories": boolean, // Whether to filter memories, default is false
  "org_id": string, // Organization ID filter
  "project_id": string // Project ID filter
}

Environment Variables

# Required settings
MEM0_API_KEY=your_api_key # Required for cloud mode

OR
OPENAI_API_KEY=your_api_key # Required for local mode

# Optional settings
DEFAULT_USER_ID=user123 # Default user ID
PORT=3000 # Custom service port
LOG_LEVEL=info # Log level

Security Considerations

• Ensure the security of API keys and avoid leakage.
• Use encryption technology when handling sensitive data.
• Regularly back up important memory data.

Error Handling

Performance Optimization

• Regularly clean up expired memories.
• Configure an appropriate caching strategy.
• Use load balancing to handle high - concurrency requests.

Monitoring and Logging

• Integrate application performance monitoring tools.
• Analyze log files to troubleshoot problems.
• Set up an alert mechanism to respond to abnormal situations in a timely manner.

Troubleshooting

If you encounter problems, please follow these steps:

1. Check if the service is running normally.
2. View the latest logs to obtain error information.
3. Confirm that the API key and configuration are correct.
4. Contact the Mem0 support team for help.

Appendix

• Supported Filter Operators: =, >, <, >=, <=, !=
• Log Format: [timestamp] [level] [message]
• Default Port: 3000

⚠️ Important Note

The content of this document may be updated. It is recommended to regularly consult the official Mem0 documentation for the latest information.