manhattan-mcp 0.1.18

MCP Server for Manhattan Memory System - Give AI agents persistent memory

Manhattan MCP

Give AI Agents Persistent Memory - MCP Server for the Manhattan Memory System

Manhattan MCP is a local Model Context Protocol (MCP) server that connects AI agents (Claude Desktop, Cursor, Windsurf, etc.) to the Manhattan Memory System - a cloud-based persistent memory for AI assistants.

Features

• 🧠 Persistent Memory - Store and retrieve information across conversations
• 🔍 Semantic Search - Find relevant memories using natural language queries
• 🤖 AI-Generated Answers - Get comprehensive answers using memory context
• 👤 Multi-Agent Support - Create separate memory spaces for different use cases
• 📊 Analytics - Track memory usage and agent statistics
• 💾 Export/Import - Backup and restore memory data

Installation

pip install manhattan-mcp

Quick Start

1. Get Your API Key

Sign up at https://themanhattanproject.ai to get your API key.

2. Set Environment Variable

export MANHATTAN_API_KEY="your-api-key-here"

Or create a .env file:

MANHATTAN_API_KEY=your-api-key-here

3. Configure Your AI Client

Claude Desktop

Add to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "manhattan": {
      "command": "manhattan-mcp",
      "args": ["start"]
    }
  }
}

Cursor

Add to your Cursor MCP settings (.cursor/mcp.json):

{
  "mcpServers": {
    "manhattan": {
      "command": "manhattan-mcp"
    }
  }
}

Windsurf

Add to your Windsurf configuration:

{
  "mcpServers": {
    "manhattan": {
      "command": "manhattan-mcp",
      "args": ["start"]
    }
  }
}

4. Start Using Memory!

Once configured, your AI agent will have access to 35+ memory tools:

• search_memory - Search for relevant memories
• add_memory_direct - Store new information
• get_context_answer - Get AI-generated answers with memory context
• session_start / session_end - Manage conversation sessions
• And many more!

Available Tools

Memory Operations

Tool	Description
search_memory	Search memories using natural language
add_memory_direct	Store structured memories
get_context_answer	Get AI answers using memory context
update_memory_entry	Update existing memories
delete_memory_entries	Delete specific memories

Agent Management

Tool	Description
create_agent	Create a new memory agent
list_agents	List all your agents
get_agent	Get agent details
update_agent	Update agent configuration
delete_agent	Permanently delete an agent

Session Management

Tool	Description
session_start	Initialize a conversation
session_end	End session and sync memories
pull_context	Load relevant context
push_memories	Sync pending memories

AI Helpers

Tool	Description
auto_remember	Automatically extract facts from messages
should_remember	Check if info is worth storing
what_do_i_know	Summary of known user info

Configuration Options

Environment Variable	Description	Default
MANHATTAN_API_KEY	Your API key (required)	-
MANHATTAN_API_URL	API endpoint URL	https://themanhattanproject.ai/mcp
MANHATTAN_AGENT_ID	Default agent ID	Enterprise default
MANHATTAN_TIMEOUT	Request timeout (seconds)	120

CLI Commands

# Start the MCP server (default)
manhattan-mcp start

# Show version
manhattan-mcp --version

# Show help
manhattan-mcp --help

Example Usage

Once your AI agent is connected, it can use memory like this:

Storing information:

User: My name is Sarah and I prefer Python over JavaScript.
AI: *calls add_memory_direct to store this preference*
    Nice to meet you, Sarah! I've noted your preference for Python.

Retrieving context:

User: What programming language should I use for this project?
AI: *calls search_memory to find preferences*
    Based on your preference for Python, I'd recommend using it for this project!

Development

# Clone the repository
git clone https://github.com/agent-architects/manhattan-mcp
cd manhattan-mcp

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

License

MIT License - see LICENSE for details.

Links

• 🌐 Website
• 📖 Documentation
• 🐛 Issues
• 💬 Discord