N8N

You run an MCP server that lets you build and manipulate n8n workflows in Cursor IDE by prompting an AI agent. It creates valid n8n workflow JSON you can copy into the n8n workflow editor, enabling programmatic workflow creation, updates, and AI-assisted wiring.

How to use

You interact with the n8n Workflow Builder MCP server through your MCP client. Start the server in your Cursor project, then configure your MCP client to point at the server. The server exposes tools to create, modify, and validate n8n workflows, discover available node types, and wire connections between nodes. Use the AI-oriented interfaces to compose complex AI-driven workflows by linking model, memory, embeddings, tools, triggers, and AI agents.

How to install

Prerequisites you need on your machine before you begin:

• Node.js v14 or newer
• Cursor IDE v0.48 or newer
• npm (for npm i or npx usage)

Choose one of the MCP server installation approaches below and follow the steps exactly.

Using npx in mcp.json (Easiest)

{
  "mcpServers": {
    "n8n-workflow-builder": {
      "command": "npx",
      "args": [
        "-y",
        "n8n-workflow-builder-mcp"
      ],
      "env": {
        "N8N_API_URL": "http://localhost:5678",
        "N8N_API_KEY": "your-n8n-api-key-here",
        // "N8N_VERSION": "1.76.1" // optional, if not set, the server will try to detect the version automatically
      }
    }
  }
}

Setup steps

1. Create the .cursor directory at the project root if it does not exist: mkdir -p .cursor

2. Create or update the configuration file at .cursor/mcp.json with the block above, replacing N8N_API_URL and N8N_API_KEY with your values. You may also specify N8N_VERSION to override automatic detection.

3. Restart Cursor IDE for changes to take effect.

Alternative: Development installation

For development or local testing, you can clone and build from source.

git clone https://github.com/ifmelate/n8n-workflow-builder-mcp.git
cd n8n-workflow-builder-mcp
npm install
npm run build

Dev mode with auto-rebuild

npm run dev

Cursor IDE integration

You can configure the MCP server through Cursor settings if you prefer a UI-based setup. Enable the server, then provide the runtime command and environment variables as shown in the npx example above.

Available MCP tools and how to use them

The server provides a suite of tools to manage and compose n8n workflows. You can create and list workflows, add and modify nodes, manage connections (including AI-aware connections), plan and apply workflow changes, and compose AI-driven workflows in one call.

Configuration and security notes

• The MCP server uses the n8n API, so you must provide a valid N8N_API_URL and N8N_API_KEY. Keep the API key secure and do not expose it in client-side code. - If you run locally, ensure your n8n instance is reachable at the configured URL. - You can optionally specify N8N_VERSION to help with compatibility when auto-detect is not desired.

Troubleshooting

• If the MCP server is not visible in Cursor, verify your .cursor/mcp.json formatting and that Cursor has been restarted after changes. - Check the MCP-specific logs in Cursor’s Output panel to isolate issues with the server start or tool registration. - If you encounter network or version issues, ensure your Node.js and npm versions are up to date and try the development installation route for debugging.

Notes

The server is designed to interact with AI agents for building and wiring n8n workflows. It supports automatic version detection for a wide range of n8n versions and filters nodes to maintain compatibility.

Available tools

create_workflow

Create a new n8n workflow with a given name and workspace context.

list_workflows

List workflows in the workspace with optional limits and cursors.

get_workflow_details

Retrieve detailed information about a specific workflow.

validate_workflow

Validate a workflow against node schemas and connectivity.

add_node

Add a new node to a workflow with optional position and parameters.

edit_node

Edit an existing node within a workflow.

delete_node

Remove a node from a workflow.

list_available_nodes

List available node types with optional filtering and version matching.

add_connection

Create a connection between two nodes in a workflow.

add_ai_connections

Wire AI components (model, memory, embeddings, tools) to an agent in a workflow.

connect_main_chain

Build a minimal main path through AI workflow nodes to ensure connectivity.

plan_workflow

Create a non-destructive plan for updating a workflow.

review_workflow_plan

Validate a plan in-memory and report errors, warnings, and suggestions.

apply_workflow_plan

Apply a reviewed plan to the on-disk workflow atomically.

compose_ai_workflow

Create a complex AI-driven workflow including wiring and basic validation.

suggest_node_params

Suggest minimal valid parameters for a node type using defaults.

list_missing_parameters

List required parameters that are missing for a node.

fix_node_params

Fill in defaults for missing required node parameters.

list_template_examples

Show usage examples for node templates and templates by type.

get_n8n_version_info

Return information about the current n8n version and capabilities.