Check environment is valid and determine safety level:

• test/staging: Standard teardown (1 confirmation)
• prod: Production teardown (3 confirmations, typed confirmation)

Step 2: Load Configuration

Source cloud-common config loader:

source plugins/faber-cloud/skills/cloud-common/scripts/config-loader.sh
load_config

Extract:

• Terraform directory path
• AWS profile for environment
• State backup location

Step 3: Backup Terraform State

Execute backup script:

./plugins/faber-cloud/skills/infra-teardown/scripts/backup-state.sh $ENV

Creates timestamped backup:

• Location: infrastructure/backups/terraform-state-{env}-{timestamp}.tfstate
• Verifies backup created successfully

Step 4: Confirmation(s)

Non-Production (test, staging):

If --confirm flag NOT present:

• Show resources to be destroyed (terraform plan -destroy)
• Show estimated cost savings
• Request 1 confirmation: "Destroy {count} resources in {env}? (yes/no)"

If --confirm flag present:

• Skip confirmation, proceed directly

Production:

IGNORE --confirm flag (reject with error if provided)

Require 3 separate confirmations:

1. Initial confirmation: "You are about to destroy PRODUCTION infrastructure. This cannot be undone. Proceed? (yes/no)"
2. Plan review: Show terraform plan -destroy output, request review confirmation
3. Typed confirmation: User must type environment name exactly: "Type 'prod' to confirm destruction:"

Between confirmations, allow user to cancel at any point.

Step 5: Execute Pre-Destroy Hooks

Execute pre-destroy hooks:

bash plugins/faber-cloud/skills/cloud-common/scripts/execute-hooks.sh pre-destroy {environment} {terraform_dir}

CRITICAL:

• If pre-destroy hooks fail (exit code 1): STOP teardown, show error
• If pre-destroy hooks pass (exit code 0): Continue to Step 6
• Pre-destroy hooks are essential for production safety (backups, notifications, etc.)

Step 6: Execute Destruction

Execute destroy script:

./plugins/faber-cloud/skills/infra-teardown/scripts/destroy.sh $ENV

This script:

• Sets appropriate timeout (10 min for non-prod, 30 min for prod)
• Executes: terraform destroy -auto-approve
• Captures output
• Returns exit code

Step 7: Execute Post-Destroy Hooks

Execute post-destroy hooks:

bash plugins/faber-cloud/skills/cloud-common/scripts/execute-hooks.sh post-destroy {environment} {terraform_dir}

• If post-destroy hooks fail: WARN user, destruction complete but post-destroy actions failed
• If post-destroy hooks pass: Continue to Step 8

Step 8: Verify Removal

Execute verification script:

./plugins/faber-cloud/skills/infra-teardown/scripts/verify-removal.sh $ENV

This script:

• Checks Terraform state is empty
• Queries AWS to verify resources removed
• Returns list of any remaining resources (should be empty)

Step 9: Document Teardown

Execute documentation script:

./plugins/faber-cloud/skills/infra-teardown/scripts/document-teardown.sh $ENV

Appends to deployment history (docs/infrastructure/deployments.md):

## Teardown - {env} - {timestamp}

**Destroyed by:** {user}
**Reason:** {reason or "Manual teardown"}
**Resources removed:** {count}
**Cost savings:** ${monthly_cost}/month
**State backup:** infrastructure/backups/terraform-state-{env}-{timestamp}.tfstate

### Resources Destroyed:
- {resource_type}: {resource_name}
- ...

Step 10: Report Results

Output summary:

✅ Infrastructure Teardown Complete

Environment: {env}
Resources destroyed: {count}
State backup: infrastructure/backups/terraform-state-{env}-{timestamp}.tfstate
Cost savings: ${monthly_cost}/month

All resources verified removed from AWS.

Deployment history updated: docs/infrastructure/deployments.md