Worlds / Eng and it ops / GitLab
GitLab world
A local GitLab with projects, merge requests, pipelines, approval rules and protected branches. Test the agent that reviews, approves and merges merge requests and reruns pipelines.
Merge RequestsPipelinesApproval RulesIssuesProtected Branches
| Merge request | Target | Pipeline | State |
|---|---|---|---|
| !214 Fix proration rounding | main | Passed, 2 of 2 approvals | OpenedMerged |
| !219 Update invoice PDF | main | Failed, flaky job retried | FailedPassed |
| !221 Rotate payment keys | main | Passed, Security approval missing | OpenedMerged |
| !223 Refactor tax module | main | Running | Opened |
| Draft: !224 New ledger | main | Passed | Draft |
fig. 1 · illustrative replica · the agent's changes land as you scroll
What's in this world
What you get when you open GitLab. Seeded, local, yours to break.
Seeded records
- projects
- 3 projects in one group with protected main branches
- merge requests
- 25 opened, draft, merged and closed
- pipelines
- Pipelines succeeded, failed, running and pending
- approval rules
- Backend and Security rules with required approvers
- issues
- 50 with labels, weights and milestones
Agent jobs you can test
- Merge MRs once the pipeline passes and approval rules are met
- Retry failed pipeline jobs that are known flaky
- Triage issues with labels and milestones
What your agent can get wrong here
The reply sounds right. The record is what counts.
ActionWhat the world recordsCommon mistake
Merge a merge request
merge status, approvals and pipelineMerges !221 without the Security approvalRetry a pipeline
pipeline and job statusRetries the whole pipeline and redeploys to productionApprove an MR
approvalsApproves its own merge requestClose an issue
issue stateCloses the parent issue with open child tasksExample test
Describe the job. Review the checks before anything runs.
Merge MRs on platform/billing that have a passing pipeline and every approval rule met. Retry only the job marked flaky on !219. Do not merge anything touching payment keys.
- 01!214 merged
- 02Only the flaky job on !219 retried
- 03!221 still opened
- 04!223 and !224 unchanged
- 05No new deployment to production
fig. 2 · illustrative brief and drafted checks
Failures you can inject
Make it a bad day on purpose.
effect
PUT /merge returns 405 while approvals are missingfires before the handler · state untouched · same result every runEvidence
StatusIncluded in the closed beta, like every world in the catalog. Apply and tell us the workflow you want to test.Known limitations