Description

This project is in PREVIEW and has not been officially released, try it out and create issues

n8n-nodes-letta

This is the official n8n node that allows you to integrate Letta AI agents into your n8n workflows.

n8n is a fair-code licensed workflow automation platform.

Letta is a platform for building stateful AI agents with long-term memory.

Quick Start

New to this node? Check out our 5-minute Quick Start Guide and demo workflows to get started quickly!

Developers: See DEVELOPMENT.md for local testing and development setup.

Installation

Community Node (Recommended)

1. Go to Settings > Community Nodes in your n8n instance
2. Select Install
3. Enter @letta-ai/n8n-nodes-letta in Enter npm package name
4. Agree to the risks of using community nodes
5. Select Install

Manual Installation

To get started install the package in your n8n root directory:

npm install @letta-ai/n8n-nodes-letta

For Docker-based n8n installations, add the following line before the $N8N_CUSTOM_EXTENSIONS variable in your .env file:

N8N_CUSTOM_EXTENSIONS=@letta-ai/n8n-nodes-letta

Configuration

Credentials

This node requires Letta API credentials:

1. Go to your Letta dashboard
2. Navigate to API settings and generate an API token
3. In n8n, create new Letta API credentials by clicking on the "+"
4. Select "Letta API" as the credential type and click on "Continue".
5. Fill your credential details:
   • API Token: Your Letta API token
   • Base URL: https://api.letta.com (default) or your self-hosted instance URL

Operations

Send Message

Send a message to a Letta agent and receive a response.

Parameters

Required:

• Agent ID: The ID of your Letta agent (e.g., agent_abc123)
• Role: The role of the message sender
  • user: Message from the user
  • system: System message
  • assistant: Message from the assistant
• Message: The message content to send

Optional:

• Max Steps: Maximum number of steps the agent can take (1-100, default: 10)
• Use Assistant Message: Include assistant message in response (default: true)
• Enable Thinking: Enable agent's thinking process in response (default: false)
• Return Message Types: Filter message types in response
  • Internal Monologue
  • Function Call
  • Function Return
  • Reasoning

Output

The node returns the complete response from the Letta API, including:

• messages: Array of messages from the agent
• stop_reason: Reason for ending the conversation
• usage: Token and step usage statistics

Demo Workflows

We provide ready-to-use example workflows in the demo/ directory:

1. Simple Chat – Basic message sending and response handling
2. Webhook Chat API – REST API endpoint for chat applications
3. Scheduled Summary – Automated daily summaries via email

📚 View Full Demo Documentation with detailed setup instructions, usage examples, and best practices.

Quick Example: Simple Chat

┌─────────┐     ┌──────────┐     ┌─────────┐
│ Trigger │────▶│  Letta   │────▶│ Display │
└─────────┘     └──────────┘     └─────────┘

1. Manual Trigger: Click to test
2. Letta Node: Sends message to agent
3. Display: Shows agent response

Import this workflow

API Reference

This node uses the Letta Messages API:

• Endpoint: POST /v1/agents/{agent_id}/messages
• Documentation: https://docs.letta.com/api-reference/agents/messages/create

Compatibility

• n8n version: 1.0.0+
• Letta API version: v1

Resources

• n8n community nodes documentation
• Letta documentation
• Letta API reference

Development

Setup

# Install dependencies
npm install

# Build the node
npm run build

# Watch for changes during development
npm run dev

# Lint code
npm run lint

# Format code
npm run format

Project Structure

letta-n8n-node/
├── credentials/
│   └── LettaApi.credentials.ts    # API credentials definition
├── nodes/
│   └── Letta/
│       ├── Letta.node.ts          # Main node implementation
│       └── letta.svg              # Node icon
├── dist/                          # Compiled JavaScript (generated)
├── package.json
├── tsconfig.json
└── README.md

Testing in n8n

To test your changes locally:

1. Build the node: npm run build
2. Link to your n8n installation: npm link
3. In your n8n directory: npm link @letta-ai/n8n-nodes-letta
4. Restart n8n

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT

Support

For issues and questions:

• Node issues: Open an issue in this repository
• n8n questions: n8n community forum
• Letta questions: Letta documentation