Automated test screenshots: diagnose failed scenarios
A failed scenario points to something worth investigating, but the result alone may not explain the cause. Start with run history and the affected step. Combine the action, the unmet expectation and the page appearance at the time of the test.
Check the run state first
Queued means waiting for execution; running means the scenario is being executed. Neither is a verdict. The list refreshes while runs are pending. Completed run details show timing, the result, its reason and retained steps.
Inconclusive confirms neither a healthy store nor an application failure. It may indicate an unavailable browser, a verification screen or an inability to perform the check. An assertion failure means the observed state did not satisfy a specific condition.
How do you open a step screenshot?
- Open the scenario history and select a run.
- Find the step you want to inspect.
- Expand the step screenshot and open the full image if you need a larger view.
Screenshots are available for browser scenarios. Enable screenshot saving in the scenario editor to capture successful steps. An image shows the visible browser viewport at capture time, not a recording of the entire run. Deep checks do not launch a browser or produce these images.
Why did a passing test produce no image?
A successful test and a saved screenshot are separate outcomes. First check whether capture was enabled when the run started. Capture or file storage may also have failed. A storage problem does not turn a working page into a failed test. Some failed runs cannot have screenshots, for example if the browser never started.
Details of ordinary successful runs may be sampled. Enabling screenshots for a browser scenario retains the associated steps between samples as well. Images are private, available through the panel and subject to run retention.
Narrow down the cause
- Missing text or element: compare the expectation with page content and language, and check the selector.
- Timeout after a click: check whether the test incorrectly expects navigation after a JavaScript update.
- Verification screen: review WAF rules and the test identification information in the panel. DockRay does not solve captchas.
- Application error: inspect exceptions reported in the same time window. Correlation can guide diagnosis but does not prove causation.
Related errors require your application to report them to the same DockRay project. After a fix, rerun the scenario and compare the condition that previously failed.
See a repeatable shopping cart scenario and learn about testing after deployment.