How to Set Up n8n MCP Server (2026)

The Model Context Protocol (MCP) allows AI agents like Claude Code, Cursor, and Cline to connect to external tools and services. By setting up n8n as an MCP server, you can give your AI assistant access to hundreds of integrations -- from Slack and Google Sheets to databases and custom APIs -- all through n8n's visual workflow builder.

This guide walks you through setting up an n8n MCP server from scratch, connecting it to AI tools, and building practical automation workflows.

What Is MCP and Why n8n?

Model Context Protocol (MCP)

MCP is an open standard (created by Anthropic) that defines how AI models connect to external tools. Instead of hardcoding integrations, MCP lets you expose tools as a standardized interface that any compatible AI agent can use.

Why n8n for MCP?

n8n is a workflow automation platform (similar to Zapier but self-hostable and open source) with 400+ built-in integrations. By exposing n8n workflows as MCP tools, your AI agent gains access to:

Category	Example Integrations
Communication	Slack, Discord, Email, Telegram
Productivity	Google Sheets, Notion, Airtable, Todoist
Development	GitHub, GitLab, Jira, Linear
Databases	PostgreSQL, MySQL, MongoDB, Supabase
Cloud	AWS, GCP, Azure, Cloudflare
Finance	Stripe, QuickBooks, PayPal
CRM	Salesforce, HubSpot, Pipedrive
Custom	Any REST API, webhooks, scripts

Prerequisites

Docker and Docker Compose installed
Node.js 18+ for the MCP bridge
An AI tool that supports MCP (Claude Code, Cursor, Cline, etc.)
Basic familiarity with n8n (optional, we cover the basics)

Step 1: Deploy n8n

The easiest way to run n8n is with Docker:

# Create a directory for n8n data
mkdir -p ~/n8n-data

# Run n8n with Docker
docker run -d \
  --name n8n \
  -p 5678:5678 \
  -v ~/n8n-data:/home/node/.n8n \
  -e N8N_SECURE_COOKIE=false \
  -e WEBHOOK_URL=http://localhost:5678/ \
  n8nio/n8n:latest

Or use Docker Compose for a more robust setup:

# docker-compose.yml
version: '3.8'
services:
  n8n:
    image: n8nio/n8n:latest
    ports:
      - "5678:5678"
    environment:
      - N8N_SECURE_COOKIE=false
      - WEBHOOK_URL=http://localhost:5678/
      - N8N_BASIC_AUTH_ACTIVE=true
      - N8N_BASIC_AUTH_USER=admin
      - N8N_BASIC_AUTH_PASSWORD=changeme
    volumes:
      - n8n_data:/home/node/.n8n
    restart: unless-stopped

volumes:
  n8n_data:

docker compose up -d

Open http://localhost:5678 to access the n8n editor.

Step 2: Create n8n Workflows for MCP

Each n8n workflow that you want to expose as an MCP tool needs a Webhook trigger. This is how the MCP bridge will invoke the workflow.

Example 1: Slack Message Sender

In n8n, click Add Workflow
Add a Webhook node as the trigger:
- Method: POST
- Path: /mcp/send-slack-message
Add a Slack node:
- Action: Send Message
- Channel: {{ $json.channel }}
- Message: {{ $json.message }}
Activate the workflow

Test it with curl:

curl -X POST http://localhost:5678/webhook/mcp/send-slack-message \
  -H "Content-Type: application/json" \
  -d '{"channel": "#general", "message": "Hello from MCP!"}'

Example 2: Google Sheets Lookup

Create a new workflow with a Webhook trigger:
- Method: POST
- Path: /mcp/lookup-spreadsheet
Add a Google Sheets node:
- Action: Read Rows
- Spreadsheet ID: your-spreadsheet-id
- Range: Sheet1!A:Z
Add a Filter node to match the query
Add a Respond to Webhook node to return results
Activate the workflow

Example 3: Database Query

Webhook (POST /mcp/query-database)
  → PostgreSQL (Execute Query)
  → Respond to Webhook (Return Results)

Step 3: Install the n8n MCP Bridge

The MCP bridge translates between the MCP protocol and n8n webhook calls. Install the community bridge package:

npm install -g @anthropic-ai/mcp-server-n8n

Or use the community package:

npm install -g n8n-mcp-server

Step 4: Configure the MCP Server

Create a configuration file that maps n8n workflows to MCP tools:

{
  "n8n_base_url": "http://localhost:5678",
  "n8n_api_key": "your-n8n-api-key",
  "tools": [
    {
      "name": "send_slack_message",
      "description": "Send a message to a Slack channel",
      "webhook_path": "/webhook/mcp/send-slack-message",
      "parameters": {
        "channel": {
          "type": "string",
          "description": "Slack channel name (e.g., #general)",
          "required": true
        },
        "message": {
          "type": "string",
          "description": "Message text to send",
          "required": true
        }
      }
    },
    {
      "name": "lookup_spreadsheet",
      "description": "Look up data in the team Google Spreadsheet",
      "webhook_path": "/webhook/mcp/lookup-spreadsheet",
      "parameters": {
        "query": {
          "type": "string",
          "description": "Search term to look up",
          "required": true
        }
      }
    },
    {
      "name": "query_database",
      "description": "Run a read-only SQL query against the project database",
      "webhook_path": "/webhook/mcp/query-database",
      "parameters": {
        "sql": {
          "type": "string",
          "description": "SQL SELECT query to execute",
          "required": true
        }
      }
    },
    {
      "name": "create_github_issue",
      "description": "Create a GitHub issue in the project repository",
      "webhook_path": "/webhook/mcp/create-github-issue",
      "parameters": {
        "title": {
          "type": "string",
          "description": "Issue title",
          "required": true
        },
        "body": {
          "type": "string",
          "description": "Issue description in markdown",
          "required": true
        },
        "labels": {
          "type": "string",
          "description": "Comma-separated labels",
          "required": false
        }
      }
    }
  ]
}

Save this as ~/.config/n8n-mcp/config.json.

Step 5: Connect to Claude Code

Add the n8n MCP server to Claude Code's MCP configuration.

Edit ~/.claude/mcp_servers.json:

{
  "mcpServers": {
    "n8n": {
      "command": "n8n-mcp-server",
      "args": ["--config", "/Users/yourname/.config/n8n-mcp/config.json"]
    }
  }
}

Restart Claude Code:

claude

Claude Code now has access to your n8n tools. Test it:

You: Send a Slack message to #dev-updates saying "Deployment complete for v2.1.0"

Claude Code will call the send_slack_message tool, which triggers the n8n workflow, which sends the Slack message.

Step 6: Connect to Cursor or Cline

Cursor MCP Configuration

In Cursor, go to Settings > MCP Servers and add:

{
  "n8n": {
    "command": "n8n-mcp-server",
    "args": ["--config", "/Users/yourname/.config/n8n-mcp/config.json"]
  }
}

Cline MCP Configuration

In Cline's settings panel, add the MCP server:

{
  "mcpServers": {
    "n8n": {
      "command": "n8n-mcp-server",
      "args": ["--config", "/Users/yourname/.config/n8n-mcp/config.json"]
    }
  }
}

Practical Use Cases

Use Case 1: AI-Powered Deployment Notifications

When Claude Code finishes a task, it can notify your team:

You: Fix the pagination bug in src/api/users.ts, commit the fix,
     and send a Slack message to #dev-updates with a summary.

Claude Code will:

Fix the code
Create a Git commit
Call the send_slack_message MCP tool to notify the team

Use Case 2: Data-Driven Development

You: Look up the top 10 customers by revenue in the spreadsheet
     and create a dashboard component that displays them.

Claude Code will:

Call lookup_spreadsheet to get the data
Generate a React component using the real data structure

Use Case 3: Issue Triage Automation

You: Review the error logs, identify the root cause, fix it,
     and create a GitHub issue documenting what happened.

Use Case 4: Database-Informed Coding

You: Query the database for the schema of the users table
     and generate a TypeScript interface that matches it.

Building More Complex n8n Workflows

Multi-Step Workflow: Bug Report Pipeline

Webhook (POST /mcp/report-bug)
  → Create GitHub Issue
  → Create Jira Ticket
  → Send Slack Notification to #bugs
  → Log to Google Sheet
  → Respond to Webhook (Return Issue URL)

This single MCP tool triggers a complete bug reporting pipeline across four services.

Scheduled Workflows with MCP Triggers

Combine MCP-triggered workflows with n8n's scheduling:

n8n Cron (Daily at 9am)
  → Query Database for Errors
  → If errors > threshold
    → Send Alert via Slack
    → Create Summary Report

Security Best Practices

Practice	Implementation
Use API keys	Set `N8N_BASIC_AUTH_ACTIVE=true` and use credentials
Limit SQL queries	Only allow SELECT statements in database tools
Validate inputs	Add validation nodes in n8n workflows
Use HTTPS	Put n8n behind a reverse proxy with TLS
Restrict network	Run n8n on an internal network, not publicly exposed
Audit logging	Enable n8n execution logging for all MCP-triggered workflows

Troubleshooting

Issue	Solution
"MCP server not found"	Verify the server is installed: `which n8n-mcp-server`
"Connection refused"	Check that n8n is running: `docker ps`
"Webhook not found"	Ensure the workflow is activated in n8n
"Authentication failed"	Verify API key in the MCP config matches n8n
Tool not appearing	Restart Claude Code / Cursor after config changes
Timeout errors	Increase timeout in n8n webhook node settings

Conclusion

Setting up n8n as an MCP server transforms your AI coding assistant into a powerful automation hub. Instead of manually switching between tools, your AI agent can send Slack messages, query databases, create issues, and trigger complex workflows -- all from natural language commands.

If your n8n automation workflows involve AI media generation tasks like creating images, videos, or audio, Hypereal AI provides REST APIs that integrate perfectly with n8n's HTTP request nodes. Build end-to-end AI media pipelines that trigger from your MCP-connected AI assistant.