mcp-n8n-workflow-builder

This MCP server provides an AI-powered interface to create, manage, and monitor n8n workflows using natural language. It exposes a suite of 17 tools that cover workflow lifecycle operations, execution monitoring, tagging, and credential management, enabling you to describe your desired automation in plain English and have the system generate production-ready n8n configurations. You can access the server via Claude AI or Cursor IDE integrations, which let you issue natural language commands like creating a webhook workflow, updating an existing workflow, or inspecting recent runs, without manually editing JSON every time. The server maintains multi-instance support, so you can work with development, staging, and production environments from a single MCP instance, with internal routing to the appropriate n8n environment.

To use it, install the MCP server client (via npx as demonstrated in the Quick Start) and connect it to your Claude Desktop or Cursor IDE configuration. Then start issuing conversational prompts such as: create a new webhook workflow in production that triggers on HTTP POST /customer-signup, validates fields, sends a Slack notification, and stores results in PostgreSQL; or list active workflows in staging and check their last execution times. The integrated tools handle creation, updating, activation, execution, monitoring, retrying, analysis of runs, and credential management, so you can complete end-to-end automation workflows without leaving your IDE or chat interface.

Prerequisites:

• Node.js v14+ (v18+ recommended)
• npm v7+ (for installation and npm-based workflows)
• Access to an n8n instance with API capabilities

Installation steps:

1. Install the MCP server client globally using npm:

npm install -g @kernel.salacoste/n8n-workflow-builder

2. Verify the installation:

npx @kernel.salacoste/n8n-workflow-builder --version

3. Start or configure the MCP server according to your environment. If using npx, you can invoke the server as demonstrated in the Quick Start section of the README. Ensure your n8n instance is reachable from the environment running the MCP server and that any required API keys are available to the MCP process.

4. Optional integration setup ( Claude Desktop / Cursor IDE )

• Claude Desktop: add the MCP server entry to claude_desktop_config.json (as shown in the README) and restart Claude Desktop.
• Cursor IDE: add the MCP server entry to .cursor/mcp.json in your workspace and reload Cursor.

Prerequisites summary:

• Node.js and npm installed
• Access to an n8n instance (with API access)
• Claude Desktop or Cursor IDE for natural-language control (optional but recommended)

Tips and common considerations:

• The MCP server relies on npx to fetch and run the npm package @kernel.salacoste/n8n-workflow-builder. An active internet connection is required during the first run to install dependencies.
• If you are behind a firewall, ensure outbound HTTP(S) access is allowed for npm registry and any API endpoints used by your n8n instance.
• For multi-environment setups, keep environment-specific credentials (n8n_host and n8n_api_key) consistent and securely stored in your environment files or config.json, as described in the Quick Start configuration.
• When using Claude Desktop, you may need to restart the app after updating .config.json or claude_desktop_config.json to ensure the MCP server is discoverable.
• The MCP server supports credential management and API key encryption features; avoid exposing sensitive data in logs by adjusting log level and ensuring proper credential handling in your n8n instances.