Results

Is the build done?

curl -u "x:$TESTKASE_ACCESS_KEY" https://hub.testkase.com/espresso/v1/builds/<build_id>

{
  "build_id": "5b1e7c2a-9d4f-4e1b-8a63-2f0c7d9e4b18",
  "status": "failed",
  "devices": [
    { "device": "Pixel 8-17", "sessionId": "0747163a-…", "status": "failed",
      "tests": { "total": 3, "passed": 1, "failed": 1, "skipped": 1 } }
  ]
}

Poll until status is no longer queued or running. The build is passed only when every device passed, so gate your CI on status == "passed".

Which tests failed, and why

curl -u "x:$TESTKASE_ACCESS_KEY" https://hub.testkase.com/espresso/v1/builds/<build_id>/sessions/<sessionId>

You get every test with its status, the stack trace of each failed test, a videoUrl per test, and links to the device log, test output and crash dump. The links work for 30 minutes, so ask again for fresh ones.

If the app crashes, the test that was running fails, the remaining tests on that device do not run, and the device ends as error.

Stop a build

curl -u "x:$TESTKASE_ACCESS_KEY" -X POST https://hub.testkase.com/espresso/v1/builds/<build_id>/stop

Running and waiting devices end as stopped. It is safe to call on a finished build.

In the dashboard

Every build is listed in Automate → Mobile Apps → Runs under its buildTag. Each device's session there has a video per test, the All Commands tab (each Espresso action and check, like an Appium session) and the logs.

Billing works like Appium: each device is billed per minute. Time waiting in the queue is free.

Full response fields: build status, one device, stop.