lab-terraform

Terraform monorepo that manages every AWS resource the lab provisions.

• Source of truth: this GitHub repo.
• What applies the changes: HCP Terraform (Terraform Cloud), connected to this repo as a VCS-driven workspace.
• What lab members do: edit .tf files in a branch, open a PR, get the speculative plan reviewed, merge.
• What no one needs: the Terraform CLI on their laptop. The university-managed OS doesn't allow it, and the workflow doesn't need it.

Start here

Read deliverable.md. It's the lab-staff handbook — how the repo is organized, what to do day-to-day, how to start a new client project, and how to transplant a project to a client at delivery.

Visual workflow diagrams live in docs/flowchart.md.

Repo at a glance

modules/              # Reusable building blocks (s3-static-site, cognito-user-pool, …)
projects/<client>/    # Per-client project deployments — each folder = one HCP workspace
docs/                 # Handbook, flow diagrams, conventions, transplant procedure
.github/workflows/    # CI checks (fmt + validate + tflint) — no apply, HCP does that

Conventions in one line

• Every project takes the same five variables: client_name, project_name, environment, aws_region, tags.
• Resource names follow ${client_name}-${project_name}-${environment}-<resource>.
• No hardcoded AWS account IDs, ARNs, or lab-specific resource names anywhere. Everything must work in someone else's AWS account on day one — that's what makes transplant cheap.

Full conventions: docs/conventions.md.