| name | infra-release-readiness |
| description | Operator-grade workflow for PetSwipe deployment, Kubernetes, Terraform, release bundle, and production-readiness checks. |
| disable-model-invocation | true |
Use this skill when you explicitly want deployment-readiness or infra/devops work.
Workflow
- Inspect the touched infra area first: Kubernetes, Terraform, Docker Compose, scripts, or docs.
- Run the strongest repo-native validation available.
- Separate what is repo-ready from what still depends on real operator values.
- Update operator docs when the workflow changes.
- Do not claim enterprise readiness unless the repo and operator prerequisites both support it.
Repo-Specific Guidance
- Kubernetes validation lives in
make k8s-preflight, make k8s-render, and make k8s-render-prod.
- Terraform validation starts with
make tf-preflight ENV=production.
- Deploy/release guardrails exist in:
scripts/deploy-preflight.sh
scripts/terraform-preflight.sh
scripts/k8s-preflight.sh
scripts/release-bundle.sh
- Production still depends on real operator inputs like
.env.production, terraform/backend.hcl, terraform/environments/*.tfvars, and non-placeholder Kubernetes values.
Use the infra-deployment-auditor subagent for isolated review when the task is large.
@references/deployment-checklist.md