Troubleshoot test failures

  • Release version: Australia
  • Updated March 12, 2026
  • 1 minute to read
  • Use ATF troubleshooting agent to significantly reduce the skills and resources needed to troubleshoot test failures on covered metadata.

    Before you begin

    Note:
    You must have either of these roles to access the ATF module and test results tables.
    • system_admin
    • atf_test_admin
    • atf_test_designer
    • atf_ws_designer
    Role required: atf_triager + now_assist_panel_user

    Procedure

    1. Navigate to All > Automated Test Framework (ATF) > Tests.
    2. Select the test you want to run.
      Note:
      You can also select New if you want to create a new test.
    3. Optional: On the Test new record form, enter all the details to create the test.
      Note:
      This step is applicable only if you have created a new test in the previous step.
    4. Optional: Add new test steps to the existing or newly created test.
    5. Select Run Test.
      The Run Test modal shows up.
    6. Select Go to Result to view the information on the test execution.
      If you instead select the Test Results related list after the test execution is completed, a list of step summary shows up. You are required to select a step summary record with a Failure status to know more about the reasons of its failure.
      The Test Results form shows up.
      Note:
      You can access the Test Results form by selecting Go to Result on the Run Test modal or from the Test Results related list.
    7. Select Triage with Otto on the Test Results form.
      The ServiceNow IDE or ServiceNow Studio interface shows up.
    8. Review the prompt that is pre-populated depending on the test you're running and the selected summary record with a Failure status.
    9. Follow the automatic triaging process for the failed record.
      When a test fails, the agent automatically starts the troubleshooting process. It identifies the step where the failure occurred and analyzes the root cause. The agent then recommends possible fixes. After you select a recommended option, the agent applies the fix, updates the test code, and reruns the test automatically. When the test passes, it provides a summary of the changes made to resolve the failure.
      Note:
      See Author, execute, and troubleshoot tests with Test Agent if you want to execute a new application or a test.