About n8n Workflow Builder MCP
n8n Workflow Builder MCP is an MCP server for creating, inspecting, editing, and validating n8n workflow files locally. It is for AI-agent users who want to generate workflows for import or version-aware use without a running n8n instance; execution is not implemented.
Key features
- 01Create and edit workflow files
- 02Discover and filter n8n nodes
- 03Connect regular and AI nodes
- 04Plan, review, and apply changes
- 05Validate workflow structure
- 06Match node versions to n8n releases
Use cases
- 01Generate workflows offline for later import
- 02Prepare workflows for older self-hosted n8n
- 03Inspect and validate workflow connections
MCP support
It exposes workflow, node, connection, planning, parameter, and validation tools to MCP-compatible clients such as Claude Code, VS Code, and Cursor. It writes workflow JSON locally and does not execute workflows.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEn8n Workflow Builder MCP
A Model Context Protocol (MCP) server for building and manipulating n8n workflows. Build n8n workflows just by prompting with AI — works with Claude Code, VS Code, Cursor, and any MCP-compatible client.
Is this the right MCP server for you?
This server writes n8n workflow JSON files locally. It doesn't need a running n8n instance, and it picks node versions from a bundled catalog that covers n8n 1.86.0 – 2.6.2.
Use this server if you:
- want to generate workflow files offline and import them yourself (or commit them to git)
- are on an older self-hosted n8n and need workflows whose node
typeVersions your instance actually supports. SetN8N_VERSIONto match it.
Consider an alternative if you:
- want the AI to build, test and run workflows directly in your n8n instance. Use n8n's built-in instance-level MCP server.
- want a large node documentation and template library with deployment through the n8n API. Use czlonkowski/n8n-mcp.
- need nodes added in n8n after 2.6.2. They aren't in the bundled catalog yet.
DEMO VIDEO:
Cursor rules
- File with rules is in
rules/n8n-mcp-server-rules.mdc
Key Features
- Workflow Management: Create, update, and execute n8n workflows programmatically (execute is not implemented yet)
- Node Discovery: Explore available n8n nodes and their capabilities
- Connection Management: Create connections between workflow nodes
- AI Integration: Special tools for connecting AI components in workflows
- AI-Friendly Interface: Designed specifically for interaction with AI agents
- N8N Version Management: Automatic version detection and compatibility handling - supports 184+ n8n versions (1.86.0 – 2.6.2) with dynamic node filtering and "closest lower version" matching for backward compatibility
Prerequisites
- Node.js (v18 or higher)
- npm (for npx command)
- An MCP-compatible client (Claude Code, VS Code, Cursor, etc.)
Installation & Setup
Getting your n8n API Key
The API key is optional. The server only uses N8N_API_URL/N8N_API_KEY to detect your n8n version. To work fully offline, skip them and set N8N_VERSION (e.g. N8N_VERSION=1.123.2) instead.
- Open your n8n instance in a browser
- Go to Settings > API Keys
- Click Create API Key
- Copy the generated key and use it in your configuration
Claude Code (Recommended)
Add the MCP server using the Claude Code CLI:
claude mcp add n8n-workflow-builder -- npx -y n8n-workflow-builder-mcp
Then set the environment variables:
claude mcp add n8n-workflow-builder \
-e N8N_API_URL=http://localhost:5678 \
-e N8N_API_KEY=your-n8n-api-key-here \
-- npx -y n8n-workflow-builder-mcp
N8N_VERSIONis optional — the server auto-detects it from the API.
VS Code / Cursor
Add to your MCP config file (.vscode/mcp.json for VS Code, .cursor/mcp.json for Cursor):
{
"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"
}
}
}
}
Restart your IDE for changes to take effect.
Development Installation
For development or local testing, 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
Then point your MCP client to the built entry point:
# Claude Code
claude mcp add n8n-workflow-builder -- node /absolute/path/to/n8n-workflow-builder-mcp/dist/index.js
# VS Code / Cursor — use the same JSON config above with "command": "node" and "args": ["/absolute/path/to/dist/index.js"]
For development with auto-rebuild:
npm run dev
Available MCP Tools
The server provides the following tools for working with n8n workflows:
Core Workflow Management
| Tool Name | Description | Key Parameters |
|---|---|---|
| create_workflow | Create a new n8n workflow | workflow_name, workspace_dir |
| list_workflows | List workflows in the workspace | limit (optional), cursor (optional) |
| get_workflow_details | Get detailed information about a specific workflow | workflow_name, workflow_path (optional) |
| validate_workflow | Validate a workflow file against node schemas and connectivity | workflow_name, workflow_path (optional) |
Node Management
| Tool Name | Description | Key Parameters |
|---|---|---|
| add_node | Add a new node to a workflow | workflow_name, node_type, position (optional), parameters (optional), node_name (optional), typeVersion (optional), webhookId (optional), workflow_path (optional), connect_from (optional), connect_to (optional) |
| edit_node | Edit an existing node in a workflow | workflow_name, node_id, node_type (optional), node_name (optional), position (optional), parameters (optional), typeVersion (optional), webhookId (optional), workflow_path (optional), connect_from (optional), connect_to (optional) |
| delete_node | Delete a node from a workflow | workflow_name, node_id, workflow_path (optional) |
| list_available_nodes | List available node types with optional filtering. Supports tag-style synonyms and multi-token OR/AND logic | search_term (optional), n8n_version (optional), limit (optional), cursor (optional), tags (optional, default: true), token_logic (optional: 'or' default, or 'and') |
Connection Management
| Tool Name | Description | Key Parameters |
|---|---|---|
| add_connection | Create a connection between two nodes | workflow_name, source_node_id, source_node_output_name, target_node_id, target_node_input_name, target_node_input_index (optional), workflow_path (optional) |
| add_ai_connections | Wire AI model, tools, and memory to an agent | workflow_name, agent_node_id, model_node_id (optional), tool_node_ids (optional), memory_node_id (optional), embeddings_node_id (optional), vector_store_node_id (optional), vector_insert_node_id (optional), vector_tool_node_id (optional), workflow_path (optional) |
| connect_main_chain | Build a minimal main path through AI workflow nodes (Trigger → Model → Memory → Embeddings → Doc Loader → Vector Store → Vector Tool → Agent) | workflow_name, workflow_path (optional), dry_run (optional), idempotency_key (optional) |
Workflow Planning & Composition
| Tool Name | Description | Key Parameters |
|---|---|---|
| plan_workflow | Create a non-destructive plan (nodes and connections) to update a workflow. Does not write files | workflow_name, target (nodes, connections), workspace_dir (optional) |
| review_workflow_plan | Apply a plan in-memory and return validation errors, warnings, and suggested fixes. Does not write files | workflow_name, plan, workflow_path (optional) |
| apply_workflow_plan | Apply a previously reviewed plan to the workflow on disk (atomic write) | workflow_name, plan, workflow_path (optional) |
| compose_ai_workflow | Compose a complex AI workflow (agent + model + memory + embeddings + vector + tools + trigger) in one call, including wiring and basic validation | workflow_name, plan, n8n_version (optional) |
Parameter Management
| Tool Name | Description | Key Parameters |
|---|---|---|
| suggest_node_params | Suggest minimal valid parameters for a node type using defaults and required fields | node_type, typeVersion (optional), existing_parameters (optional) |
| list_missing_parameters | List required parameters missing for a node considering visibility rules | node_type, typeVersion (optional), parameters |
| fix_node_params | Return parameters with defaults applied for required fields that are missing | node_type, typeVersion (optional), parameters (optional) |
Templates & Discovery
| Tool Name | Description | Key Parameters |
|---|---|---|
| list_template_examples | List node usage examples extracted from free templates. Filter by node_type or template_name | node_type (optional), template_name (optional), limit (optional), cursor (optional) |
| get_n8n_version_info | Get current N8N version and capabilities | random_string |
Validation behavior
validate_workflow promotes warnings to errors and additionally fails when any enabled node is not connected (directly or via AI ports) to the main chain starting at the inferred startNode. Use connect_from/connect_to or add_ai_connections to fix connectivity.
Troubleshooting
General
- Check your MCP config — make sure JSON is valid and the server name matches.
- Update Node.js to the latest LTS version.
- Clear npm cache if npx fails:
npm cache clean --force - Try global install as a fallback:
npm install -g n8n-workflow-builder-mcp
Claude Code
- Run
claude mcp listto verify the server is registered. - Check logs with
claude mcp logs n8n-workflow-builder.
VS Code / Cursor
- Check the Output panel — select "MCP" from the dropdown to see server logs.
- Make sure the server is enabled in Settings > Features > MCP Servers.
- Restart the IDE after config changes.
Project Structure
/src: Main source code/src/tools: MCP tools implementation/src/models: Data models/src/utils: Utility functions/src/middleware: Authentication and middleware/config: Configuration files/tests: Test files/workflow_nodes: n8n node definitions/docs: Additional documentation
Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
License
MIT License
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.