How to Use Terraform MCP with Claude Code

What is the Terraform MCP?

The Terraform MCP integrates Claude Code with the Terraform CLI and Terraform Cloud/HCP Terraform APIs. Claude can generate HCL modules, run terraform plan, inspect state files, validate configurations, and manage workspaces — making infrastructure-as-code workflows dramatically faster and less error-prone.

Installation

mcpizy install terraform

The MCP requires Terraform CLI to be installed on your machine. For Terraform Cloud integration, provide a TFC organization and API token during setup.

Key Capabilities

• Generate Terraform modules — describe infrastructure in plain English; Claude writes the HCL.
• Run terraform plan — execute plans and parse the output to summarize what will change.
• Inspect state — list resources in state and show their current attributes.
• Validate configurations — run terraform validate and explain any errors.
• Manage workspaces — create and switch between Terraform Cloud workspaces.

Example Usage

Provision a new S3 bucket with versioning and a lifecycle rule:

resource "aws_s3_bucket" "artifacts" {
  bucket = "my-app-artifacts-prod"
}

resource "aws_s3_bucket_versioning" "artifacts" {
  bucket = aws_s3_bucket.artifacts.id
  versioning_configuration { status = "Enabled" }
}

resource "aws_s3_bucket_lifecycle_configuration" "artifacts" {
  bucket = aws_s3_bucket.artifacts.id
  rule {
    id     = "expire-old"
    status = "Enabled"
    expiration { days = 90 }
  }
}

Claude generates this HCL, runs the plan, and reports back the expected changes.

Tips and Best Practices

• Always review the plan output before asking Claude to apply — never auto-apply in production.
• Store sensitive variables in Terraform Cloud variable sets rather than in .tfvars files.
• Combine with Kubernetes MCP to manage both cluster infrastructure and application workloads from one session.
• Ask Claude to add resource tagging to all AWS/GCP/Azure resources for cost attribution.

View on the MCPizy directory