Before changing a workspace from Terraform 1.13 to 1.15, the team wants to see how the configuration behaves on 1.15 without altering the workspace's version setting and without any risk of an apply. What should it do?
Select an answer to reveal the explanation.
Short Explanation
A speculative plan is a dress rehearsal: you can't apply it, and when you retry a plan-only run you get to pick the Terraform version. You see 1.15's behaviour without touching the workspace setting.
Full Explanation
Speculative plans are plan-only runs that HCP Terraform creates for pull requests and that can never be applied. When retrying a plan-only run, HCP Terraform lets the user choose a different Terraform version for that retry, and the plan details note that it used a version other than the workspace's setting. That shows exactly how the configuration plans on 1.15 while the workspace stays on 1.13 and nothing can be applied. HashiCorp's upgrade tutorial uses this pattern before changing the workspace version. Changing the workspace setting and discarding the run alters the very setting the team wants to leave alone, and any other run queued meanwhile, including one from a merge, would execute on 1.15. A refresh-only plan updates state to match real infrastructure; it does not select a different Terraform release, and it still runs with the workspace's configured version. terraform version only prints the local binary's version and provider versions; it does not analyse a remote workspace's configuration. Exam caveat: a clean speculative plan does not prove every edge case, so a non-production workspace on the new version adds confidence before prod. Operational check: open the retried run, confirm it reports 1.15 and differs from the workspace version, and review the plan for unexpected changes.