Full lifecycle
Plan, apply, drift detection, and rollback — real Terraform and cloud API calls.
Plan, apply, drift detection, and rollback aren't decorative statuses — they run real Terraform and cloud API calls against your actual infrastructure state.
- Plan computes real changes before anything deploys, with cost and credential preflight checks
- Apply actually provisions infrastructure — it isn't a progress bar
- Drift detection continuously compares live state against declared state
- Rollback restores a prior state in reverse-topological order when something breaks
How it works
Plan and apply run as queued workflow steps — init, plan, a policy gate, then apply — with results streamed back live. A separate scheduler compares live infrastructure against declared state on a recurring basis, and rollback restores a prior state per-backend (S3, GCS, Azure RM, Terraform Cloud, HTTP, or local), in reverse-topological order.
Rollback runs in reverse-topological order specifically so dependent resources aren't torn down before what depends on them — and each state backend (S3, GCS, Azure RM, Terraform Cloud) has its own rollback implementation, not one generic code path pretending they're the same.