Some stages run E2E tests — end-to-end checks against your deployed application — as part of clearing the stage. A stage can run those tests automatically, or it can wait for you to start them. This page covers the manual case: starting the tests yourself and reading the result. For what a stage and its gate are, see The pipeline: stages and gates.
When you trigger E2E tests yourself
Each stage has an E2E trigger setting in its configuration under Deliver > Pipeline, offered as three buttons:
- None — the stage runs no E2E tests; once its pull request is merged it advances straight on.
- Automatic — DIJJI.ai starts the tests for you once the stage’s deployment completes. It needs at least one E2E target set on the stage (see below); without one, the tests have nothing to run against.
- Manual — the run waits, and you start the tests yourself when your deployment is ready.
This page covers the Manual case. With Manual, you control when the new code reaches the environment the tests run against — so you deploy first, then run the tests against what you deployed.
Set the stage’s E2E targets
A stage that runs E2E tests — Automatic or Manual — needs to know what to test. You set that up ahead of time on the pipeline, not at trigger time. Open Deliver > Pipeline, and on the stage’s configuration find the E2E Targets list (it appears once the stage’s E2E trigger is set to Automatic or Manual).
Select Add target to add a target, then fill in:
- Title — a label for the target, like Web App, Backoffice, or API.
- Check type — choose Browser for end-to-end checks against a deployed application, or API for checks against a deployed service.
- URL — the deployed address this target’s checks run against.
Each target can also carry a Test Username and Test Password so the checks can sign in and exercise screens behind a login. These are optional — leave them blank to skip the login for that target. Use a throwaway test account only, never a password you use elsewhere.
Add as many targets as the stage needs, then save the pipeline. The Trigger E2E Tests dialog reads this list, so what you set here is exactly what runs.
Find the Trigger E2E button
Each stage the run has reached appears as a card — both on the task’s detail page and on the run detail page. On a stage that’s waiting for its E2E tests — after the stage’s pull request is merged — a Trigger E2E button appears on the stage’s card.
The button is there when the tests haven’t run yet, and it stays available after a failed attempt so you can try again.
Confirm the deploy, then run
Selecting Trigger E2E opens the Trigger E2E Tests dialog. The tests “will run against the targets configured for this stage” — so the dialog lists the targets you set on the pipeline (each shows its title, check type, and URL) and there’s nothing to type here.
If the stage has no targets yet, the dialog warns “No targets configured for this stage. Add them in the pipeline settings first.” Close the dialog, go to Deliver > Pipeline, add at least one E2E target to the stage (see above), save, then trigger again.
Below the target list is a confirmation checkbox — “Has the new code been deployed to these URLs?” — and you can’t start the tests until it’s ticked.
That checkbox matters. The tests run against whatever is currently at those URLs. As the dialog warns, “even if the PR is merged, E2E tests the old code unless it has been physically deployed” — a false positive that tells you nothing. Deploy first, tick the box, then start.
With the box ticked, select Run E2E Tests to start.
Skip the tests instead
If a stage is waiting on its E2E tests but you don’t need them to run — a trivial change, or you’ve verified the deployment another way — a Skip E2E button sits next to Trigger E2E on the stage card. Selecting it asks you to confirm “Skip E2E?”: the stage is marked complete without any test verification and the run advances to the next stage. Use it deliberately — skipping records no test result for that stage.
Read the result
The stage card shows an E2E status badge while the tests run and after they finish:
- Running — the tests are in progress.
- Passed — the tests succeeded.
- Failed — the tests found a problem.
A details line and a full E2E Details section show the test output where it’s available, and a Previous E2E Attempts list keeps every run of the tests so you can compare attempts.
If the tests fail, the Trigger E2E button stays on the card. Fix what the output points to — redeploy if needed — and trigger again against the stage’s targets. If a target itself changed, update it in Deliver > Pipeline first. If they pass, the stage has what it needs and the run carries on.
Triggering E2E tests isn’t admin-only — anyone on the project’s team who can open the run can do it.
For every state a run can pass through while it waits on you, see Run states.