MCP Infrastructure as Code Assistant

An MCP server for managing infrastructure as code with Terraform.

Features

• Initialize Terraform working directories

• Generate and show execution plans

• Apply changes to infrastructure

• Destroy infrastructure

• Validate Terraform configurations

• Show current state or saved plans

• Manage Terraform workspaces

Related MCP server: Terry-Form MCP

Prerequisites

• Python 3.8 or higher

• Terraform 1.5.7 or higher

• Docker and Docker Compose (optional)

Installation

Local Installation

1. Clone the repository:

   git clone https://github.com/yourusername/mcp-iac.git
   cd mcp-iac

2. Install dependencies using uv:

   curl -LsSf https://astral.sh/uv/install.sh | sh
   uv pip install -e .

Docker Installation

1. Clone the repository:

   git clone https://github.com/yourusername/mcp-iac.git
   cd mcp-iac

2. Build and run the Docker container:

   docker-compose up -d

Usage

Local Usage

1. Start the MCP server:

   python main.py

2. Use the MCP CLI to interact with the server:

   mcp terraform_init --working-dir ./terraform
   mcp terraform_plan --working-dir ./terraform
   mcp terraform_apply --working-dir ./terraform --auto-approve

Docker Usage

1. Start the MCP server:

   docker-compose up -d

2. Use the MCP CLI to interact with the server:

   mcp terraform_init --working-dir ./terraform
   mcp terraform_plan --working-dir ./terraform
   mcp terraform_apply --working-dir ./terraform --auto-approve

Example Terraform Configuration

The repository includes an example Terraform configuration that creates an EC2 instance in AWS:

terraform {
  required_providers {
    aws = {
      source  = "hashicorp/aws"
      version = "~> 5.0"
    }
  }
}

provider "aws" {
  region = var.region
}

resource "aws_instance" "example" {
  ami           = var.ami_id
  instance_type = var.instance_type

  tags = {
    Name = var.instance_name
  }
}

Contributing

1. Fork the repository

2. Create a feature branch

3. Commit your changes

4. Push to the branch

5. Create a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Available Tools

• terraform_init: Initialize a Terraform working directory

• terraform_plan: Generate and show an execution plan for Terraform

• terraform_apply: Apply the changes required to reach the desired state

• terraform_destroy: Destroy the infrastructure managed by Terraform

• terraform_validate: Validate the syntax and internal consistency of Terraform files

• terraform_show: Show the current state or a saved plan

• terraform_workspace_list: List Terraform workspaces

• terraform_workspace_select: Select a Terraform workspace

Example Usage

Here's an example of how to use the MCP server with an AI agent:

1. Start the MCP server:

   python main.py

2. Connect to the server using an MCP client:

   mcp connect http://localhost:8000

3. The AI agent can now help you with Terraform operations. For example:

   • Initialize a Terraform working directory

   • Generate and review execution plans

   • Apply changes to infrastructure

   • Destroy infrastructure resources

   • Validate Terraform configurations

Examples

Check out the examples directory for sample Terraform configurations that demonstrate how to use the MCP server:

• examples/aws-s3: A simple AWS S3 bucket example