| name | tf-plan |
| description | Planning orchestrator for Terraform infrastructure provisioning. Drives spec-driven development from requirements intake through specification, clarification, planning, review, tasks, and analysis.Entry point for the planning workflow. |
Planning Orchestrator
Workflow Steps
- Execute the phases sequentially. Do not move ahead until the step previous has completed.
- Before and after each subagent run: post progress to the gh issue.
- use the following template for gh issues:
bash .foundations/scripts/bash/post-issue-progress.sh $ISSUE_NUMBER "<step>" "<status>" "<summary>" "$DETAILS"
example
bash .foundations/scripts/bash/post-issue-progress.sh $ISSUE_NUMBER "Research" "started"
bash .foundations/scripts/bash/post-issue-progress.sh $ISSUE_NUMBER "Research" "complete" $DETAILS
Phase 1 — Setup
- Run validate-env.sh
- Check for error and stop if pre-requisites not met
- Then call MCP
list_terraform_orgs to verify TFE_TOKEN.
- Gather requirements from the user using
.github/ISSUE_TEMPLATE/terraform-agent-provisioning.yml as a guide. Cover required fields first, then ask about optional sections (network, monitoring, backup/DR, cost, tags)
Phase 2 — Specification
- run sdd-specify subagent with the captured requirements, creates spec.md
- run sdd-checklist subagent against
spec.md
- run sdd-clarify subagent — pass
checklist directory findings as input alongside spec.md. Use AskUserQuestion for any HIGH-impact gaps not covered during requirements intake
Phase 3 — Research + Planning
- run mutiple concurrent sdd-research subagents (parallel)
- run mutiple concurrent sdd-plan-draft subagents to produce, creates
plan.md, data-model.md, contracts/module-interfaces.md
Phase 4 — Tasks + Analysis
- Run mutiple concurrent sdd-tasks subagents, creates
tasks.md
- Run mutiple concurrent sdd-analyze subagents in parallel. pass in
tasks.md createsanalysis.md
Phase 5 — Summary + Approval
- Compile results, post to GitHub issue, add agent:awaiting-review label.
Display: > Planning is complete. Please review the artifacts in specs/<branch>/ and approve before proceeding to implementation. Run /tf-implement when ready.
