🤖 AI Agents Integration¶

EasyPanel MCP works with any AI agent that supports the Model Context Protocol (MCP). This guide covers integration with popular AI agents and IDEs.

📋 Supported AI Agents¶

Agent	Type	Setup Complexity	Best For
Claude Desktop	Desktop App	⭐ Easy	General purpose, natural language
Cursor IDE	Code Editor	⭐ Easy	Development workflows
Cline	VS Code Extension	⭐ Easy	Coding assistance
ChatGPT Desktop	Desktop App	⭐⭐ Medium	GPT-4 infrastructure management
n8n	Automation Platform	⭐⭐ Medium	Workflow automation
Custom Agents	Any MCP Client	⭐⭐⭐ Advanced	Custom integrations

🎯 Claude Desktop Integration¶

What is Claude Desktop?¶

Claude Desktop is Anthropic's desktop application that supports MCP servers, allowing Claude to interact with external tools and services.

Installation¶

Download Claude Desktop
macOS: Download from Anthropic
Windows: Coming soon
Linux: Use Claude via web or third-party clients

Install on macOS

# Install to Applications
mv ~/Downloads/Claude.app /Applications/

Configuration¶

Open Claude Desktop config file:

# macOS
open ~/Library/Application\ Support/Claude/claude_desktop_config.json

# Windows (when available)
%APPDATA%\Claude\claude_desktop_config.json

Add EasyPanel MCP configuration:

{
  "mcpServers": {
    "easypanel": {
      "command": "python",
      "args": ["/path/to/easypanel-mcp/src/server.py"],
      "cwd": "/path/to/easypanel-mcp",
      "env": {
        "EASYPANEL_URL": "https://your-easypanel.com",
        "EASYPANEL_API_KEY": "ep_live_xxxxxxxxxxxxx"
      }
    }
  }
}

Restart Claude Desktop

Usage Examples¶

Once configured, you can ask Claude to:

User: "Show me all my projects in EasyPanel"
Claude: 📊 I can see you have 3 projects:
        - dashboard (Active)
        - test (Active)
        - wikileyes (Active)

User: "Deploy a PostgreSQL database in my test project"
Claude: 🚀 Creating PostgreSQL service...
        ✅ Service 'postgres-db' deployed successfully
        📦 Image: postgres:15
        🔗 Connected to project 'test'

💻 Cursor IDE Integration¶

What is Cursor?¶

Cursor is an AI-powered code editor with built-in MCP support.

Installation¶

Download Cursor
Visit: https://cursor.sh
Download for your platform (Windows, macOS, Linux)
Install and open Cursor

Configuration¶

Open Cursor Settings
Cmd/Ctrl + Shift + P → "Cursor Settings"
Or: File → Settings → AI → MCP Servers
Add EasyPanel MCP:

{
  "mcpServers": {
    "easypanel": {
      "command": "python",
      "args": ["/absolute/path/to/easypanel-mcp/src/server.py"],
      "cwd": "/absolute/path/to/easypanel-mcp",
      "env": {
        "EASYPANEL_URL": "https://your-easypanel.com",
        "EASYPANEL_API_KEY": "ep_live_xxxxxxxxxxxxx"
      }
    }
  }
}

Reload Cursor

Usage in Cursor¶

Use Cmd/Ctrl + K to open AI chat
Ask about your EasyPanel infrastructure
Deploy services directly from code comments

// @easypanel Deploy this API to production
async function deploy() {
  // Cursor will use EasyPanel MCP to deploy
}

🔧 Cline (VS Code Extension)¶

What is Cline?¶

Cline is a VS Code extension that brings AI coding assistance with MCP support.

Installation¶

Install Cline in VS Code
Open VS Code
Extensions: Ctrl+Shift+X
Search for "Cline"
Install
Configure Cline

Open Cline settings (Ctrl+,) and add MCP configuration:

{
  "cline.mcpServers": {
    "easypanel": {
      "command": "python",
      "args": ["/path/to/easypanel-mcp/src/server.py"],
      "cwd": "/path/to/easypanel-mcp",
      "env": {
        "EASYPANEL_URL": "https://your-easypanel.com",
        "EASYPANEL_API_KEY": "ep_live_xxxxxxxxxxxxx"
      }
    }
  }
}

Usage¶

Open Cline chat in VS Code
Ask Cline to manage your EasyPanel infrastructure
Deploy services while coding

🤖 ChatGPT Desktop Integration¶

What is ChatGPT Desktop?¶

Third-party desktop clients for ChatGPT that support MCP.

Supported Clients¶

ChatGPT Desktop (macOS/Windows)
GitHub: https://github.com/lencx/ChatGPT
Supports MCP via plugins
OpenAI Desktop
Check for MCP support in settings

Configuration¶

Configuration varies by client. Look for MCP or plugin settings and add EasyPanel MCP server details.

⚡ n8n Automation Integration¶

What is n8n?¶

n8n is a workflow automation platform that can integrate with MCP servers.

Installation¶

Self-host n8n:

docker run -d \
  -p 5678:5678 \
  -v n8n_data:/home/node/.n8n \
  n8nio/n8n

Or use n8n Cloud:
Visit: https://n8n.io/cloud

Configuration¶

Install MCP Node for n8n
Open n8n
Settings → Community Nodes → Install
Search for "MCP"
Add EasyPanel MCP Node:
Create new workflow
Add MCP node

Configure:

Server URL: http://localhost:8080
Transport: HTTP

Run EasyPanel MCP in HTTP mode:
```
python src/server.py http
```

Example Workflow¶

Trigger (Webhook)
    ↓
MCP Node: list_projects
    ↓
IF: Project exists?
    ↓ YES → MCP Node: deploy_service
    ↓ NO  → MCP Node: create_project → deploy_service
    ↓
Response (Success message)

🔌 Custom MCP Clients¶

HTTP Mode Configuration¶

For custom clients, run EasyPanel MCP in HTTP mode:

python src/server.py http

Default endpoint: http://127.0.0.1:8080

MCP Protocol Endpoints¶

Endpoint	Method	Description
`/mcp`	POST	Main MCP endpoint
`/`	POST	Alternative endpoint

Example Request¶

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "initialize",
  "params": {
    "protocolVersion": "2024-11-05",
    "capabilities": {},
    "clientInfo": {
      "name": "my-custom-client",
      "version": "1.0.0"
    }
  }
}

Example Response¶

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "protocolVersion": "2024-11-05",
    "capabilities": {
      "tools": {
        "list": true
      }
    },
    "serverInfo": {
      "name": "easypanel-mcp",
      "version": "1.0.0"
    }
  }
}

🛠️ Available Tools¶

Once connected, AI agents can use these tools:

Projects¶

list_projects - List all projects
get_project - Get project details
create_project - Create new project
delete_project - Delete project

Services¶

list_services - List all services
get_service - Get service details
create_service - Create new service
update_service - Update service config
delete_service - Delete service
restart_service - Restart service
deploy_service - Deploy/redeploy service

Deployments¶

list_deployments - List all deployments
get_deployment - Get deployment details
create_deployment - Create new deployment
get_deployment_logs - Get deployment logs

Networks¶

list_networks - List all networks
create_network - Create new network
delete_network - Delete network

🔒 Security Considerations¶

1. Protect API Keys¶

Never commit .env files to version control
Use environment variables or secrets managers
Rotate API keys periodically

2. Network Security¶

{
  "mcpServers": {
    "easypanel": {
      "env": {
        "MCP_HOST": "127.0.0.1",  // Only localhost access
        "MCP_PORT": "8080"
      }
    }
  }
}

3. Access Control¶

Only trusted AI agents should have MCP access
Use separate API keys for different agents
Monitor logs for unusual activity

🧪 Testing Your Integration¶

Test Connection¶

# Test MCP server
python test_connection.py

Expected output:

✅ Connected successfully!
✅ Found 3 project(s)
✅ ALL TESTS PASSED

Test with AI Agent¶

Ask your AI agent:

"List my EasyPanel projects"

If configured correctly, you should see your projects listed.

🆘 Troubleshooting¶

Agent Doesn't See MCP Tools¶

Solution: 1. Restart the AI agent application 2. Verify MCP config file syntax (JSON) 3. Check that EasyPanel MCP is running

Connection Refused¶

Solution: 1. Ensure EasyPanel MCP is running 2. Check EASYPANEL_URL is correct 3. Verify API key is valid 4. Test with python test_connection.py

Tools Not Available¶

Solution: 1. Check MCP server logs for errors 2. Verify tRPC connection to EasyPanel 3. Ensure EasyPanel is accessible

📚 Additional Resources¶

Installation Guide - Setup EasyPanel MCP
Configuration Guide - Configuration options
Tools Reference - Available tools
MCP Specification - Official MCP docs

🤖 Ready to deploy with AI? Choose your agent and start managing infrastructure with natural language!