Let Developers Provision Infrastructure Straight From Their IDE

The StackGen MCP Server is a Model Context Protocol server that connects AI assistants like Claude, Cursor, and VS Code to StackGen's infrastructure deployment platform. It enables developers to deploy, manage, and monitor cloud infrastructure using natural language directly from their preferred development environment. The MCP server acts as a bridge between your AI assistant and StackGen's infrastructure lifecycle management capabilities, allowing you to create AppStacks, provision resources, detect drift, and manage multi-cloud deployments without leaving your IDE.

StackGen MCP Server currently supports:

• Claude Desktop/Claude Code - Fully supported with native integration (easiest setup)
• Cursor - Supported via MCP extension (medium difficulty)
• Gemini CLI - Fully supported with native integration
• Windsurf - Supported via Cascade
• VS Code - Coming soon via extension

Each platform offers the same comprehensive set of infrastructure management tools, with setup difficulty varying based on the platform's native MCP support.

Setup typically takes 5-10 minutes and involves three simple steps:

1. Install StackGen and cloud2code CLIs (via Homebrew or direct download)
2. Generate a Personal Access Token from your StackGen account settings
3. Configure your AI assistant with the MCP server using either the CLI command or settings file

For Claude Code, it's as simple as running one command: claude mcp add stackgen with your authentication token. The server automatically detects CLI locations and handles configuration.

The MCP server provides 25+ tools for comprehensive infrastructure management:

• Create and manage AppStacks - Define infrastructure blueprints using architecture diagrams or natural language
• Deploy multi-cloud infrastructure - Provision resources across AWS, Azure, and GCP
• Import existing resources - Convert running cloud resources into managed IaC with Cloud2Code
• Detect and manage drift - Identify configuration differences between deployed and desired state
• Apply governance policies - Enforce security, compliance, and resource restrictions
• Sync with version control - Integrate with GitHub for CI/CD workflows

All operations are executed through conversational prompts like "Deploy my Python application to AWS ECS" or "Check if my production AppStack has any drift."

Yes, the MCP server is built with enterprise security in mind:

• Encrypted token storage - PAT tokens are stored using keychain (macOS/Linux) or encrypted file storage
• No credential logging - Tokens are never exposed in logs or error messages
• Command sanitization - All CLI commands are validated to prevent injection attacks
• TLS encryption - All external communications use secure protocols
• Process isolation - Each command runs in isolation with timeout protection
• Governance enforcement - Supports IAM restrictions, security rules, and resource policies

The server enables platform engineering teams to provide developers with AI-native infrastructure access while maintaining centralized governance and security controls.