n8n MCP Server
A Model Context Protocol (MCP) server that enables seamless management of n8n workflows directly within LLMs and AI agents through the Smithery Model Context Protocol.
Features
- List available workflows from n8n
- View workflow details
- Execute workflows
- Monitor workflow executions
- Pass parameters to workflows
- MCP-compatible interface for AI agents
Getting Started
Quick Start
Install the package
npm install @dopehunter/n8n-mcp-server
Create a .env file
cp .env.example .env
Configure your n8n connection Edit the .env file and set:
* `N8N_BASE_URL`: URL to your n8n instance (e.g., `http://localhost:5678/api`)
* `N8N_API_KEY`: Your n8n API key (generate this in n8n settings)
Start the server
npm start
Test the server
curl -X POST http://localhost:3000/mcp -H "Content-Type: application/json"
-d '{"jsonrpc":"2.0","id":"1","method":"mcp.tools.list","params":{}}'
Common Issues and Troubleshooting
- Connection Refused Errors : Make sure your n8n instance is running and accessible at the URL specified in N8N_BASE_URL
- API Key Issues : Verify your n8n API key is correct and has appropriate permissions
- Docker Issues : Ensure Docker is running before attempting to build or run the Docker image
For more detailed troubleshooting, see the Troubleshooting Guide.
Components
Tools
n8n_list_workflows
- List all workflows in the n8n instance
- Input: None
n8n_get_workflow
- Get details of a specific workflow
- Input:
workflowId (string, required): ID of the workflow to retrieve
n8n_execute_workflow
- Execute an n8n workflow
- Inputs:
workflowId (string, required): ID of the workflow to executedata (object, optional): Data to pass to the workflow
n8n_get_executions
- Get execution history for a workflow
- Inputs:
workflowId (string, required): ID of the workflow to get executions forlimit (number, optional): Maximum number of executions to return
n8n_activate_workflow
- Activate a workflow
- Input:
workflowId (string, required): ID of the workflow to activate
n8n_deactivate_workflow
- Deactivate a workflow
- Input:
workflowId (string, required): ID of the workflow to deactivate
Prerequisites
- Node.js (v14+)
- n8n instance with API access
- An LLM or AI agent that supports the Model Context Protocol
Configuration Options
Docker Configuration
{
"mcpServers": {
"n8n": {
"command": "docker",
"args": ["run", "-i", "--rm", "--init", "-e", "N8N_API_KEY=$N8N_API_KEY", "-e", "N8N_BASE_URL=$N8N_BASE_URL", "mcp/n8n-mcp-server"]
}
}
}
NPX Configuration
{
"mcpServers": {
"n8n": {
"command": "npx",
"args": ["-y", "@dopehunter/n8n-mcp-server"]
}
}
}
Installation
NPM
npm install @dopehunter/n8n-mcp-server
Direct Usage with npx
npx @dopehunter/n8n-mcp-server
From Source
git clone https://github.com/dopehunter/n8n_MCP_server_complete.git
cd n8n_MCP_server_complete
npm install
cp .env.example .env
# Edit the .env file with your n8n API details
Development
Start the development server:
npm run start:dev
Build the project:
npm run build
Run tests:
npm test
Usage With Claude or Other LLMs
Start the MCP server:
npm start
Configure your LLM client to use the MCP server:
* For Claude Desktop, use the configuration from the "Configuration Options" section.
* For other clients, point to the server URL (e.g., `http://localhost:3000/mcp`).
- Your LLM can now use n8n workflows directly through MCP commands.
Building Docker Image
docker build -t mcp/n8n-mcp-server .
API Documentation
See the API Documentation for details on the available MCP functions.
Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
License
This project is licensed under the ISC License.
