n8n-mcp Hits 19K Stars: Build n8n Automation Workflows with Natural Language via Claude — MCP Protocol Lands Another Benchmark

The Pain Point: Building Automation Workflows Is Still Too Hard

n8n is one of the most popular open-source automation workflow platforms, supporting 400+ integrations and visual node orchestration. But its core interaction mode remains manual drag-and-drop — you need to:

Know which nodes are available
Understand each node’s input/output format
Manually connect data flows between nodes
Configure parameters node by node

For non-technical users or rapid prototyping, this process remains overly cumbersome.

The Solution: MCP Protocol Bridges Claude and n8n

n8n-mcp builds a bridge between AI tools like Claude and n8n through the Model Context Protocol (MCP). The core idea is elegantly simple:

You describe the automation you want in natural language, and AI generates and deploys the n8n workflow directly.

Technical Architecture

Claude Desktop / Cursor / Windsurf
         ↓ MCP Protocol
    n8n-mcp Server
         ↓ REST API
    n8n Instance

The MCP protocol enables Claude to:

Read existing n8n workflow structure and configuration
Create entirely new workflows, including nodes, connections, and parameters
Edit specific nodes in existing workflows
Execute test runs and return results

Real-World Example

You can simply chat with Claude Desktop:

“Create an automation: whenever I receive an email with ‘invoice’ in the subject, extract the PDF attachment, OCR the amount and date, write to Google Sheets, and send a Slack notification.”

Through n8n-mcp, Claude automatically:

Creates an email trigger node (Gmail/IMAP listener)
Adds keyword filtering conditions
Adds a PDF extraction node
Connects an OCR service node
Configures a Google Sheets write node
Sets up a Slack notification node
Correctly connects all data flows

All without ever opening n8n’s visual editor.

Comparison with Similar Solutions

Solution	Interaction Mode	Learning Curve	Flexibility	n8n Support
n8n native editor	Drag-and-drop nodes	Medium	Extremely high	Native
n8n-mcp	Natural language	Low	High (AI-generated + manual fine-tuning)	Full
Claude generates JSON directly	Chat → copy-paste	Low-Medium	Medium (requires manual import)	Indirect
Other AI plugins	Partial support	Medium	Low	Limited

n8n-mcp’s core advantage is bidirectional interaction: Claude not only generates workflows but can also read and understand existing workflow logic for incremental modifications.

Project Metrics

GitHub Stars: 19,257 (264 daily increase)
Forks: 3,219
Author: @czlonkowski
License: Open Source
Supported clients: Claude Desktop, Claude Code, Windsurf, Cursor

19K+ stars indicate broad community recognition, and the 264 daily star increase shows sustained market momentum.

Getting Started

1. Install n8n

npm install n8n -g
n8n start

Or with Docker:

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  n8nio/n8n

2. Install n8n-mcp

npm install -g @czlonkowski/n8n-mcp

3. Configure Claude Desktop

Add the MCP server to your Claude Desktop configuration:

{
  "mcpServers": {
    "n8n": {
      "command": "n8n-mcp",
      "args": ["--n8n-url", "http://localhost:5678"]
    }
  }
}

4. Start Conversing

Restart Claude Desktop and simply describe the automation you want.

Cost Estimate

Component	Cost
n8n (self-hosted)	Free (open-source)
n8n-mcp	Free (open-source)
Claude Pro	$20/month
Cloud hosting (optional)	$5-20/month

Total monthly cost stays within $25-40, significantly below commercial automation tools like Zapier Pro at $49/month.

Why This Matters

n8n-mcp’s significance goes beyond a single tool — it validates a larger trend: the MCP protocol is becoming the standard channel for AI tools to integrate with external systems.

After browser operations, code editing, and file management, workflow automation is MCP’s next major landing scenario. Once AI can directly manipulate your automation infrastructure, the “describe requirements → auto-deploy” paradigm will become the new default way of working.