# Run an assessment (/docs/guides/run-an-assessment)



Run a journey only after its test plan, required actions, and selected agent environment are ready.

<Steps>
  <Step>
    ### Open a journey [#open-a-journey]

    Go to **Journeys**. Select **Run** from the list, or open the journey and choose **Run test**.
  </Step>

  <Step>
    ### Review what will run [#review-what-will-run]

    Confirm the test-case count, selected environment, expected checks, and safety limits. Return to the test plan if the scope is wrong.

    <DocImage src="/images/product/run-journey.jpg" alt="Run journey view showing what will run, safety limits, and Start test" caption="Review the test count and safety limits before starting." />
  </Step>

  <Step>
    ### Start the test [#start-the-test]

    Select **Start test**. Invokeable queues the assessment and records progress for each test case.
  </Step>

  <Step>
    ### Follow progress or leave the page [#follow-progress-or-leave-the-page]

    You can leave the page while the run continues. Open **Findings > Test results** to return to it.
  </Step>

  <Step>
    ### Wait for verification and cleanup [#wait-for-verification-and-cleanup]

    The run is not complete when the agent sends its final message. Invokeable must verify the product state and confirm any required cleanup.
  </Step>

  <Step>
    ### Review the result [#review-the-result]

    Open the completed test result. Start with the summary, then use **Activity** to inspect the sequence and limitations.
  </Step>
</Steps>

## Cancel or stop a run [#cancel-or-stop-a-run]

* Cancel work that is still queued and has not started.
* Stop active work when continuing would be unsafe or unnecessary.
* Wait for cleanup when the run changed temporary state.

A stop request is not proof that every side effect ended immediately. Check the terminal state and cleanup result.

## Common reasons a run cannot start [#common-reasons-a-run-cannot-start]

* The journey is not approved or lacks a runnable test group.
* A required action needs setup.
* No active agent supports the required capabilities.
* The product environment is not authorized for a state-changing test.
* A required verification method is unavailable.
* A workspace limit or concurrent run limit has been reached.

Continue with [Review results](/docs/guides/review-results).