| name | terraform-deploy |
| description | Deploys Terraform stacks through the full lifecycle: fmt, validate, plan, and apply.
Supports deploying a single stack by folder name or all stacks in numeric order.
Use this skill whenever the user wants to deploy, apply, plan, validate, or format
Terraform code in this project. Also trigger when the user says things like
"deploy the VPC stack", "apply all stacks", "run terraform", "plan the backend stack",
or "tf apply". Even if the user only mentions one step (like "validate"), run the
full pipeline up to that step.
|
Terraform Deploy
Deploy Terraform stacks through a controlled pipeline: fmt -> validate -> plan -> apply.
Arguments
This skill accepts an optional stack name argument:
/terraform-deploy — deploys all stacks in the terraform/ directory, in numeric prefix order/terraform-deploy <stack-folder> — deploys a single stack (e.g., /terraform-deploy 00-remote-backend-stack)
The stack folder name corresponds to a directory under terraform/ in the project root.
Pipeline Steps
Execute these steps in order. Stop immediately if any step fails — do not proceed to the next step. Show the user the error output and help them fix it.
Step 1: Discover stacks
- If a specific stack was provided, verify the directory exists at
terraform/<stack-folder>/
- If no stack was provided ("deploy all"), list all directories under
terraform/ and sort them by their numeric prefix (e.g., 00-* before 01-* before 02-*). This ordering matters because later stacks often depend on earlier ones.
Step 2: For each stack, run the pipeline
For each stack directory, run these substeps in sequence:
2a: terraform fmt -check -recursive
Run from within the stack directory. This checks if the code is properly formatted.
- If formatting issues are found, run
terraform fmt -recursive to fix them automatically, then inform the user what was reformatted.
- If already formatted, move on.
2b: terraform init
Run terraform init to initialize the stack. Check if a backend.hcl file exists in the stack directory — if so, use terraform init -backend-config=backend.hcl. If init was already done (.terraform/ exists and lock file is current), you can add -reconfigure only if needed.
2c: terraform validate
Run terraform validate to check configuration validity.
- If validation fails, show the errors and stop. Help the user understand what's wrong.
2d: terraform plan
Run terraform plan -out=tfplan to create an execution plan.
- Show the user a summary of what will be created/changed/destroyed.
- If there are no changes, inform the user and skip the apply step for this stack.
2e: Apply
Run terraform apply tfplan immediately — do not ask for confirmation, the user has already decided to deploy by invoking this skill.
- After apply completes, clean up the
tfplan file
Step 3: Summary
After all stacks have been processed, provide a brief summary:
- Which stacks were deployed successfully
- Which stacks were skipped (no changes or user declined)
- Which stacks failed (and at which step)
Important Notes
- Always
cd into the stack directory before running any terraform commands
- If deploying all stacks and one fails, ask the user if they want to continue with remaining stacks or stop
- Environment variables and AWS credentials should already be configured in the shell — do not attempt to configure them
- If
terraform init has already been run and .terraform/ exists, you can skip re-init unless the user asks for it or there are provider/module changes
- When showing plan output, highlight the resource count summary (e.g., "Plan: 3 to add, 0 to change, 1 to destroy")
