Spacelift Certification Prep: Interview Q&As (2027)
Spacelift publishes no certification on spacelift.io. Use these 32 Q&As on runs, spaces, policies, stack dependencies, modules and drift to self-assess.
Spacelift does not publish a certification program on spacelift.io, so "Spacelift certification" questions in interviews are really a structured check of platform knowledge. This guide gives you that check as 32 Q&As grouped by the areas Spacelift's own documentation uses: runs, spaces, policies, stack dependencies, the module registry and drift detection.
When we checked spacelift.io, the site listed a blog, webinars, a resource library and customer stories under learning resources, but no exam, badge or course. The only certification it mentioned was its own FedRAMP authorisation, which is a company compliance status, not a credential for engineers. If a job ad asks for "Spacelift certification", treat it as a request for hands-on experience and prepare with this self-assessment.
For the broad interview set, start with our keeper post, Spacelift engineer interview questions. This page goes deeper on the details a certification exam would normally test: exact run states, policy deprecations, space inheritance and module versioning rules.
Key takeaways
- No Spacelift certification is published on spacelift.io; prove skill with a working stack setup in Git instead.
- Spacelift has four main run types: tracked runs, proposed runs, tasks and module test cases.
- Access, initialization and task policies are deprecated; approval policies and spaces replace them, and stack dependencies mostly replace trigger policies.
- Spaces form a tree from the
rootspace; child spaces with inheritance can use their parent's resources, such as policies, contexts and worker pools. - Module versions come from
module_versionin.spacelift/config.ymland must follow strict semantic versioning. - Drift detection runs only on private worker pools and creates proposed runs; reconciliation creates tracked runs.
Self-assessment map
| Area | What you should be able to do | Questions |
|---|---|---|
| Platform basics | Name supported tools and run types | 1 to 8 |
| Runs and states | Read a run's state and explain the next step | 9 to 13 |
| Spaces and access | Design a space tree for teams | 14 to 17 |
| Policies | Pick the right policy type and attach it | 18 to 21 |
| Stack dependencies | Pass outputs and order runs | 22 to 25 |
| Module registry and drift | Version, test and share modules; handle drift | 26 to 32 |
Freshers: platform basics
1. Is there an official Spacelift certification?
Not one published on spacelift.io when we checked. The site offers a blog, webinars, a resource library and mission guides. Check the site again before an interview, and if a certification appears later, use its official objectives over this list.
2. Which infrastructure tools can a Spacelift stack run?
The stack documentation lists Terraform, OpenTofu, Terragrunt, Pulumi, AWS CloudFormation, Kubernetes and Ansible. Interviewers often check that you know it is not Terraform-only.
3. What is a stack?
A stack joins a Git repository and path, a tool such as OpenTofu, its state and settings, and the environment it runs with. Each stack manages one piece of infrastructure, and its state reflects the latest tracked run.
4. What are the four main run types?
Tracked runs plan and can apply changes to real infrastructure. Proposed runs only plan, usually for a pull request. Tasks run an arbitrary command against the stack, such as a state operation. Module test cases test a module from the registry.
5. What does a stack with state "None" mean?
The stack has no runs yet. After the first run, the stack state follows the state of its most recent tracked run.
6. What is a context?
A context is a reusable bundle of environment variables, mounted files and hooks that you attach to many stacks, so shared settings live in one place.
7. What is a worker pool?
A set of workers that execute runs. Spacelift provides public workers, and you can run private workers in your own network so runs reach private resources and credentials stay inside your environment.
8. Which policy language does Spacelift use?
Open Policy Agent's Rego. Spacelift supports Rego v0 and v1, and its docs recommend v1 for all new policies.
Runs and states
9. A tracked run shows UNCONFIRMED. What does it mean?
The plan finished and has changes, and the stack is not set to autodeploy, so the run waits for a person to confirm or discard it. After confirmation it moves on to apply.
10. What is the difference between DISCARDED, CANCELED and STOPPED?
DISCARDED means someone rejected an unconfirmed plan instead of applying it. CANCELED and STOPPED both mean the run was halted before finishing; in practice CANCELED is used for runs that had not started work yet and STOPPED for runs halted while in progress. All three appear in the run states table in Spacelift's docs, which is the place to confirm the exact transitions.
11. A run sits in QUEUED for a long time. What do you check?
Check whether workers in the stack's pool are busy or offline, whether another run on the same stack holds the lock, and whether the run waits on an upstream stack dependency. Private pool capacity is the most common cause.
12. What does PENDING_REVIEW suggest?
The run is waiting for approval under an approval policy. It moves forward once the policy's conditions, such as two reviewers, are met.
13. Name the phases a successful tracked run usually goes through.
Queued, ready, preparing, initializing, planning, then unconfirmed or straight to applying if autodeploy is on, and finished. Knowing the order helps you place an error in the right phase.
2 to 5 years: spaces and policies
14. What are spaces for?
Spaces delegate partial admin rights to limited areas of a Spacelift account. A platform team can own the root space while each product team administers its own child space.
15. Which resources belong to a space?
Stacks, policies, contexts, modules, worker pools and cloud integrations. Every resource lives in exactly one space.
16. How does inheritance between spaces work?
A child space with inheritance enabled can use resources from its parent. That lets you manage a worker pool or a VCS integration once in a parent space and share it with every team space below.
17. What is the legacy space?
Accounts have a legacy space kept for compatibility with role-based access control from before spaces existed. New designs should use the space tree instead.
18. Which policy types are deprecated, and what replaces them?
Access policies are replaced by space access control. Initialization and task policies are replaced by approval policies. Trigger policies are mostly replaced by stack dependencies. Login, approval, notification, plan, push and intent policies remain.
19. How do you attach a policy to every production stack automatically?
Give the policy a label like autoattach:production and add the production label to the stacks. The policy attaches to every stack or module with that label. autoattach:* attaches it to all of them. Login policies cannot be auto-attached.
20. Which policy decides how a Git push is handled?
The push policy. It decides whether a push or pull request event triggers a tracked run, a proposed run or nothing, which is how you ignore changes outside a stack's folder.
21. How do you attach a policy with Terraform instead of the UI?
Use the Spacelift Terraform provider's spacelift_policy_attachment resource, which links a policy to a stack or module. Managing attachments in code keeps them reviewed like any other change.
Senior: dependencies, modules and drift
22. Which runs do stack dependencies affect?
Only tracked runs. Proposed runs and tasks ignore dependencies. A downstream stack starts after its upstream stack's tracked run finishes.
23. How do you pass a VPC ID from a network stack to an app stack?
Add an output reference on the dependency. Spacelift passes the upstream output to the downstream stack as an environment variable; for Terraform, name it with the TF_VAR_ prefix so it becomes an input variable.
24. You retried a failed upstream run and it succeeded, but downstream stacks did not run. Why?
Retrying a failed run does not trigger dependent stacks, even when the retry succeeds. You need to trigger the upstream run again. If a run fails, downstream runs in that chain are cancelled.
25. Why should you not combine trigger policies with stack dependencies?
Spacelift's docs say the two should not be combined because it causes confusion and can create loops. Dependencies form a directed acyclic graph, and Spacelift rejects a dependency that would create a cycle.
26. How does the module registry name a module?
As spacelift.io/<organization>/<module_name>/<provider>. If the repository follows the terraform-<provider>-<name> convention, Spacelift infers the name and provider.
27. Where does a module version come from?
From module_version in .spacelift/config.yml. When tests pass on a tracked change, Spacelift publishes that version. It must follow strict semantic versioning, so you cannot jump from 0.2.0 to 0.4.0, and you cannot reuse a published number. A missing or misnamed file falls back to 0.1.0.
version: 1
module_version: 0.3.0
test_defaults:
terraform_version: 1.9.8
tests:
- name: default-inputs
- name: rejects-bad-cidr
negative: true
environment:
TF_VAR_cidr: "not-a-cidr"
28. What does a module test case do?
It runs init, plan, apply and destroy against the module with the inputs you give, passed as TF_VAR_ variables. A test marked negative is expected to fail. Tests run on proposed and tracked changes and each reports its own commit status.
29. How do you share a module with other teams?
In the module's availability settings, choose the spaces that can use it. Child spaces with inheritance also get access. You need admin access on the module's space and write access (or the share-module permission) on the target space.
30. What are the requirements for drift detection?
It works only on private worker pools, and the docs list it for Starter plans and above. You create a schedule per stack with one or more cron rules. Worker pools can limit or disable drift detection runs.
31. What happens when drift is found with reconciliation on?
The detection job is a proposed run. With reconciliation enabled, detected drift triggers a tracked run that follows the stack's usual rules, including autodeploy and plan policies. Without it, drifted resources still show in the Resources view and webhooks include a driftDetection field.
32. How would you prove Spacelift skill without a certificate?
Publish a small repository that creates spaces, two dependent stacks with an output reference, an auto-attached plan policy, a tested module and a drift schedule, all through the Spacelift Terraform provider. Walking an interviewer through that repo shows more than a badge would.
Study path
Read the run, space, policy and module registry pages in Spacelift's docs, then build the repository from question 32 on a trial account. The scenario-based Spacelift questions and advanced Spacelift questions add practice, and the Spacelift prep guide helps with planning. Spacelift sits on top of Terraform and OpenTofu, so strengthen those with the 120+ Terraform interview questions and the IaC tools overview.
Our Terraform training covers state, modules and pipelines, in the Kharadi, Pune classroom or live online.
FAQ
Does Spacelift offer a certification?
No certification was published on spacelift.io when we checked. Its only listed certification was the company's FedRAMP authorisation.
How should I prepare for "Spacelift certification" interview questions?
Study runs, spaces, policies, stack dependencies, the module registry and drift detection in Spacelift's docs, and build a demo repository.
What are Spacelift's main run types?
Tracked runs, proposed runs, tasks and module test cases.
Which Spacelift policies are deprecated?
Access, initialization and task policies. Trigger policies are mostly replaced by stack dependencies.
Does drift detection work on public workers?
No. Spacelift's docs say drift detection only works on private workers.
How are module versions set in Spacelift?
With module_version in .spacelift/config.yml, using strict semantic versioning.
Which tools can Spacelift run besides Terraform?
OpenTofu, Terragrunt, Pulumi, AWS CloudFormation, Kubernetes and Ansible.
References
What's Your Reaction?
Like
0
Dislike
0
Love
0
Funny
0
Angry
0
Sad
0
Wow
0