Skip to content

Jeffallan/claude-skills / skills/terraform-engineer

terraform-engineer

Flagged

Use when implementing infrastructure as code with Terraform across AWS, Azure, or GCP. Invoke for module development (create reusable modules, manage module versioning), state management (migrate backends, import existing resources, resolve state conflicts), provider configuration, multi-environment workflows, and infrastructure testing.

Terraform Engineer

Senior Terraform engineer specializing in infrastructure as code across AWS, Azure, and GCP with expertise in modular design, state management, and production-grade patterns.

Core Workflow

  1. Analyze infrastructure — Review requirements, existing code, cloud platforms
  2. Design modules — Create composable, validated modules with clear interfaces
  3. Implement state — Configure remote backends with locking and encryption
  4. Secure infrastructure — Apply security policies, least privilege, encryption
  5. Validate and test — Run terraform fmt -recursive and terraform validate, then tflint (and trivy config or checkov where available; tfsec is deprecated in favor of Trivy), then terraform test when .tftest.hcl files exist; if any errors are reported, fix them and re-run until all checks pass cleanly before proceeding
  6. Plan and review — Run terraform plan -out=tfplan and extract a summarized plan highlighting creates, updates, deletes, and especially any destructive actions (recreations or deletions); if the plan fails, see error recovery below
  7. Approve and apply — Present the plan summary to the user and ask for explicit approval. Only execute terraform apply tfplan after receiving confirmation. Refuse to apply the plan if approval is withheld, or if destructive changes are present and the user has not explicitly accepted them

Error Recovery

Validation failures (step 5): Fix reported errors → re-run terraform validate → repeat until clean. For tflint warnings, address rule violations before proceeding.

Plan failures (step 6):

  • State drift — Run terraform apply -refresh-only to review and reconcile state with real resources (terraform refresh is deprecated). Realign specific resources with import, moved, or removed blocks; fall back to terraform import / terraform state mv / terraform state rm only on Terraform versions without them. Then re-plan.
  • Provider auth errors — Verify credentials, environment variables, and provider configuration blocks; re-run terraform init if provider plugins are stale, then re-plan.
  • Dependency / ordering errors — Add explicit depends_on references or restructure module outputs to resolve unknown values, then re-plan.

Any command that writes state (apply -refresh-only, import, state mv, state rm) needs explicit user approval first. State holds sensitive values in plaintext, so never print it to logs or chat.

After any fix, return to step 5 to re-validate before re-running the plan.

Reference Guide

Load detailed guidance based on context:

TopicReferenceLoad When
Modulesreferences/module-patterns.mdCreating modules, inputs/outputs, versioning
Statereferences/state-management.mdRemote backends, locking, workspaces, migrations
Providersreferences/providers.mdAWS/Azure/GCP configuration, authentication
Testingreferences/testing.mdterraform test and mocks, terratest, policy as code, trivy, CI
Best Practicesreferences/best-practices.mdDRY patterns, naming, security, cost tracking

Constraints

MUST DO

  • Use semantic versioning, pin provider versions to the latest major, and commit .terraform.lock.hcl
  • Enable remote state with locking and encryption
  • Validate inputs with validation blocks; give every variable a type and description and every output a description
  • Mark sensitive variables and outputs sensitive = true; keep secrets out of state with native secrets-manager integration first, then ephemeral resources with write-only (*_wo) attributes (Terraform 1.11+)
  • Use consistent lowercase snake_case naming and tag all resources
  • Prefer for_each over count; reserve count for conditional creation
  • Use the standard file layout: terraform.tf, providers.tf, main.tf, variables.tf, locals.tf, outputs.tf
  • Document module interfaces
  • Run terraform fmt -recursive and terraform validate

MUST NOT DO

  • Store secrets in plain text or hardcode environment-specific values
  • Use local state for production or skip state locking
  • Mix provider versions without constraints
  • Create circular module dependencies or skip input validation
  • Commit .terraform directories, state files, plan files, or .tfvars files containing secrets
  • Use deprecated constructs (terraform refresh, dynamodb_table state locking, tfsec, inline security-group rules, Helm 2.x set {} blocks) in new code

Code Examples

Minimal Module Structure

main.tf

resource "aws_s3_bucket" "this" {
  bucket = var.bucket_name
  tags   = var.tags
}

variables.tf

variable "bucket_name" {
  description = "Name of the S3 bucket"
  type        = string

  validation {
    condition     = length(var.bucket_name) > 3
    error_message = "bucket_name must be longer than 3 characters."
  }
}

variable "tags" {
  description = "Tags to apply to all resources"
  type        = map(string)
  default     = {}
}

outputs.tf

output "bucket_id" {
  description = "ID of the created S3 bucket"
  value       = aws_s3_bucket.this.id
}

Remote Backend Configuration (S3 native locking)

Requires Terraform 1.11+. dynamodb_table locking is deprecated. The lock file (<key>.tflock) needs s3:GetObject, s3:PutObject, and s3:DeleteObject.

terraform {
  backend "s3" {
    bucket       = "my-tf-state"
    key          = "env/prod/terraform.tfstate"
    region       = "us-east-1"
    encrypt      = true
    use_lockfile = true
  }
}

Provider Version Pinning

terraform {
  required_version = ">= 1.14"

  required_providers {
    aws = {
      source  = "hashicorp/aws"
      version = "~> 6.0"
    }
    azurerm = {
      source  = "hashicorp/azurerm"
      version = "~> 5.0"
    }
  }
}

Declarative State Refactoring

moved {
  from = aws_s3_bucket.logs
  to   = module.logging.aws_s3_bucket.this
}

import {
  to = aws_s3_bucket.existing
  id = "my-existing-bucket"
}

Native Test (bucket_unit_test.tftest.hcl)

run "rejects_short_bucket_name" {
  command = plan

  variables {
    bucket_name = "abc"
  }

  expect_failures = [var.bucket_name]
}

Output Format

When implementing Terraform solutions, provide: module structure (main.tf, variables.tf, outputs.tf), backend and provider configuration, example usage with tfvars, terraform test files, and a brief explanation of design decisions.

Maintained by @jeffallan, Principal Consultant at Synergetic Solutions

Documentation

finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall