opentofu-coder

name	opentofu-coder
description	This skill guides writing Infrastructure as Code using OpenTofu (open-source Terraform fork). Use when creating .tf files, managing cloud infrastructure, configuring providers, or designing reusable modules.
allowed-tools	Read Write Edit MultiEdit Grep Glob Bash WebSearch

OpenTofu Coder

⚠️ SIMPLICITY FIRST - Default to Flat Files

ALWAYS start with the simplest approach. Only add complexity when explicitly requested.

Simple (DEFAULT) vs Overengineered

Aspect	✅ Simple (Default)	❌ Overengineered
Structure	Flat .tf files in one directory	Nested modules/ + environments/ directories
Modules	None or only remote registry modules	Custom local modules for simple resources
Environments	Workspaces OR single tfvars	Duplicate directory per environment
Variables	Inline defaults, minimal tfvars	Complex variable hierarchies
File count	3-5 .tf files total	15+ files across nested directories

When to Use Simple Approach (90% of cases)

• Managing 1-5 resources of each type
• Single provider, single region
• Small team or solo developer
• Standard infrastructure patterns

When Complexity is Justified (10% of cases)

• Enterprise multi-region, multi-account
• Reusable modules shared across teams
• Complex dependency chains
• User explicitly requests modular structure

Rule: If you can define everything in 5 flat .tf files, DO IT.

Simple Project Structure (DEFAULT)

infra/
├── main.tf           # All resources
├── variables.tf      # Input variables
├── outputs.tf        # Outputs
├── versions.tf       # Provider versions
└── terraform.tfvars  # Variable values (gitignored)

Overview

OpenTofu is a community-driven, open-source fork of Terraform under MPL-2.0 license, maintained by the Linux Foundation. It uses HashiCorp Configuration Language (HCL) for declarative infrastructure management across cloud providers.

Core Philosophy

Prioritize:

• Declarative over imperative: Describe desired state, not steps
• Idempotency: Apply safely multiple times with same result
• Modularity: Compose infrastructure from reusable modules
• State as truth: State file is the source of truth for managed resources
• Immutable infrastructure: Replace resources rather than mutate in place

HCL Syntax Essentials

Resource Blocks

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

  tags = {
    Name        = "${var.project}-web"
    Environment = var.environment
  }
}

Data Sources

data "aws_ami" "ubuntu" {
  most_recent = true

  filter {
    name   = "name"
    values = ["ubuntu/images/hvm-ssd/ubuntu-jammy-22.04-amd64-server-*"]
  }

  owners = ["099720109477"]  # Canonical
}

Variables

variable "environment" {
  description = "Deployment environment (dev, staging, prod)"
  type        = string
  default     = "dev"

  validation {
    condition     = contains(["dev", "staging", "prod"], var.environment)
    error_message = "Environment must be dev, staging, or prod."
  }
}

variable "instance_types" {
  description = "Map of environment to instance type"
  type        = map(string)
  default = {
    dev     = "t3.micro"
    staging = "t3.small"
    prod    = "t3.medium"
  }
}

Outputs

output "instance_ip" {
  description = "Public IP of the web instance"
  value       = aws_instance.web.public_ip
  sensitive   = false
}

output "database_password" {
  description = "Generated database password"
  value       = random_password.db.result
  sensitive   = true
}

Locals

locals {
  common_tags = {
    Project     = var.project
    Environment = var.environment
    ManagedBy   = "OpenTofu"
  }

  name_prefix = "${var.project}-${var.environment}"
}

Meta-Arguments

count - Create Multiple Instances

resource "aws_instance" "server" {
  count = var.server_count

  ami           = var.ami_id
  instance_type = var.instance_type

  tags = {
    Name = "${local.name_prefix}-server-${count.index}"
  }
}

for_each - Create from Map/Set

resource "aws_iam_user" "users" {
  for_each = toset(var.user_names)

  name = each.value
  path = "/users/"
}

resource "aws_security_group_rule" "ingress" {
  for_each = var.ingress_rules

  type              = "ingress"
  from_port         = each.value.port
  to_port           = each.value.port
  protocol          = each.value.protocol
  cidr_blocks       = each.value.cidr_blocks
  security_group_id = aws_security_group.main.id
}

depends_on - Explicit Dependencies

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

  depends_on = [
    aws_db_instance.database,
    aws_elasticache_cluster.cache
  ]
}

lifecycle - Control Resource Behavior

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

  lifecycle {
    prevent_destroy = true
    create_before_destroy = true
    ignore_changes = [
      tags["LastUpdated"],
      user_data
    ]
  }
}

# Replace when AMI changes
resource "aws_instance" "immutable" {
  ami           = var.ami_id
  instance_type = var.instance_type

  lifecycle {
    replace_triggered_by = [
      null_resource.ami_trigger
    ]
  }
}

Module Design

Module Structure

modules/
└── vpc/
    ├── main.tf          # Primary resources
    ├── variables.tf     # Input variables
    ├── outputs.tf       # Output values
    ├── versions.tf      # Required providers
    └── README.md        # Documentation

Calling Modules

module "vpc" {
  source = "./modules/vpc"

  cidr_block  = "10.0.0.0/16"
  environment = var.environment

  azs             = ["us-east-1a", "us-east-1b"]
  private_subnets = ["10.0.1.0/24", "10.0.2.0/24"]
  public_subnets  = ["10.0.101.0/24", "10.0.102.0/24"]
}

# Remote module with version
module "eks" {
  source  = "terraform-aws-modules/eks/aws"
  version = "~> 20.0"

  cluster_name    = local.cluster_name
  cluster_version = "1.29"

  vpc_id     = module.vpc.vpc_id
  subnet_ids = module.vpc.private_subnet_ids
}

Module Best Practices

• Expose minimal, clear interface of variables
• Use sensible defaults where possible
• Document all variables and outputs
• Avoid over-generic "god" modules
• Prefer composition over configuration flags
• Version pin remote modules

State Management

Remote Backend (S3)

terraform {
  backend "s3" {
    bucket         = "my-terraform-state"
    key            = "prod/network/terraform.tfstate"
    region         = "us-east-1"
    encrypt        = true
    dynamodb_table = "terraform-locks"
  }
}

OpenTofu State Encryption (Unique Feature)

terraform {
  encryption {
    key_provider "pbkdf2" "main" {
      passphrase = var.state_encryption_passphrase
    }

    method "aes_gcm" "encrypt" {
      keys = key_provider.pbkdf2.main
    }

    state {
      method   = method.aes_gcm.encrypt
      enforced = true
    }

    plan {
      method   = method.aes_gcm.encrypt
      enforced = true
    }
  }
}

State Commands

# List resources in state
tofu state list

# Show specific resource
tofu state show aws_instance.web

# Move resource (refactoring)
tofu state mv aws_instance.old aws_instance.new

# Remove from state (without destroying)
tofu state rm aws_instance.imported

# Import existing resource
tofu import aws_instance.web i-1234567890abcdef0

See Provider Configuration for AWS provider setup, authentication methods, and multi-provider patterns.

See Environment Strategies for workspaces and directory-based environment management.

CLI Workflow

# Initialize working directory
tofu init

# Validate configuration
tofu validate

# Format code
tofu fmt -recursive

# Preview changes
tofu plan -out=plan.tfplan

# Apply changes
tofu apply plan.tfplan

# Destroy infrastructure
tofu destroy

# Show current state
tofu show

# Refresh state from actual infrastructure
tofu refresh

Best Practices Checklist

When writing OpenTofu/Terraform code:

• Use remote backend with locking for team use
• Enable state encryption (OpenTofu feature)
• Never commit .tfstate or .tfvars with secrets to VCS
• Pin provider and module versions
• Use tofu plan before every apply
• Use lifecycle.prevent_destroy for critical resources
• Document all variables and outputs
• Use locals for computed values and tags
• Prefer for_each over count for named resources
• Use validation blocks for variable constraints
• Store secrets in secret managers, not in code

Common Patterns

Conditional Resources

resource "aws_eip" "static" {
  count = var.create_elastic_ip ? 1 : 0

  instance = aws_instance.web.id
}

Dynamic Blocks

resource "aws_security_group" "main" {
  name = "${local.name_prefix}-sg"

  dynamic "ingress" {
    for_each = var.ingress_rules
    content {
      from_port   = ingress.value.port
      to_port     = ingress.value.port
      protocol    = ingress.value.protocol
      cidr_blocks = ingress.value.cidr_blocks
    }
  }
}

References

For detailed patterns and examples:

• references/hcl-patterns.md - Advanced HCL patterns
• references/project-scaffolding.md - Directory structure, .gitignore, next_steps output, security-first variables
• references/post-provisioning.md - bin/setup-server scripts for post-infra, pre-deployment setup
• references/state-management.md - State operations and encryption
• references/provider-examples.md - Multi-cloud provider configs
• references/makefile-automation.md - Makefile workflows for plan/apply/destroy