ChatGPT MCP Server

Comprehensive Model Context Protocol (MCP) server that gives ChatGPT full access to your Windows machine with filesystem operations, command execution, and n8n workflow management.

🚀 Features

Filesystem Operations

• read_file: Read file contents
• write_file: Create or overwrite files
• list_directory: List directory contents
• create_directory: Create directories recursively
• delete_file: Delete files or directories
• search_files: Search files using glob patterns
• get_file_stats: Get file metadata
• move_file: Move or rename files

Command Execution

• execute_command: Run PowerShell commands with full output

n8n Workflow Management

• n8n_list_workflows: List all workflows
• n8n_get_workflow: Get workflow details
• n8n_activate_workflow: Activate a workflow
• n8n_deactivate_workflow: Deactivate a workflow
• n8n_execute_workflow: Execute a workflow manually
• n8n_list_executions: View workflow execution history

📦 Installation

# Navigate to the project directory
cd C:\Users\bermi\Projects\chatgpt-mcp-server

# Install dependencies
npm install

# Build the TypeScript code
npm run build

🎯 Usage

Start the Server

# Start the HTTP/SSE server
npm start

The server will start on http://localhost:3000 by default.

Connect ChatGPT

1. Open ChatGPT desktop app
2. Go to Settings → Features → MCP Servers
3. Click "New Connector"
4. Enter:
   • Name: Custom Tool (or any name)
   • Description: Full system access with filesystem and n8n
   • MCP Server URL: http://localhost:3000/sse
   • Authentication: OAuth (or None if not required)
5. Check "I understand and want to continue"
6. Click "Create"

Test the Connection

Ask ChatGPT to:

• "List files in my Projects directory"
• "Read the contents of package.json"
• "Show me all n8n workflows"
• "Execute the command 'Get-Date' in PowerShell"

🔧 Configuration

n8n API Key (Optional)

If your n8n instance requires authentication, you can either:

1. Pass it with each call - ChatGPT will include the API key in tool parameters
2. Set environment variable:

   $env:N8N_API_KEY="your-api-key-here"
   npm start
   

Change Port

$env:PORT=8080
npm start

n8n Instance

The server is configured to connect to: http://192.168.50.246:5678

To change this, edit src/index.ts line 18:

const N8N_BASE_URL = "http://your-n8n-instance:port";

🛡️ Security Considerations

⚠️ WARNING: This server gives ChatGPT full access to:

• Your entire filesystem
• Command execution on your machine
• Your n8n workflows

Recommendations:

• Only run on trusted local networks
• Review all commands before ChatGPT executes them
• Consider adding authentication/authorization
• Run with limited user permissions if possible
• Monitor all operations

📝 Development

Watch Mode

npm run watch

Project Structure

chatgpt-mcp-server/
├── src/
│   ├── index.ts          # Main MCP server (stdio-based)
│   └── http-server.ts    # HTTP/SSE wrapper for ChatGPT
├── dist/                 # Compiled JavaScript
├── package.json
├── tsconfig.json
└── README.md

🔍 Troubleshooting

ChatGPT Can't Connect

• Ensure the server is running: npm start
• Check the URL is exactly: http://localhost:3000/sse
• Verify no firewall is blocking port 3000

n8n Commands Fail

• Verify n8n is accessible: http://192.168.50.246:5678
• Check if n8n API is enabled in settings
• Ensure API key is correct (if required)

File Operations Fail

• Check file paths are correct (Windows uses backslashes)
• Verify you have permissions for the operation
• Use absolute paths when possible

📚 Examples

Read a File

ChatGPT: "Read the file at C:\Users\bermi\Projects\hourly-autopilot-system\package.json"

Create a New File

ChatGPT: "Create a new file called test.txt with the content 'Hello World'"

List n8n Workflows

ChatGPT: "Show me all my n8n workflows"

Execute a Command

ChatGPT: "Run the command 'Get-ChildItem C:\Users\bermi\Projects' in PowerShell"

Search Files

ChatGPT: "Find all TypeScript files in the hourly-autopilot-system project"

📄 License

MIT

🤝 Support

For issues or questions, please check the logs in the terminal where the server is running.