> ## Documentation Index > Fetch the complete documentation index at: https://docs.testdino.com/llms.txt > Use this file to discover all available pages before exploring further. # Run automated tests from Test Management > Pick test cases in Test Management and start a GitHub Actions run of only the automated Playwright tests linked to them. **Run automation** starts a GitHub Actions run of only the automated Playwright tests linked to the test cases you pick. You choose the test cases, a branch, and a workflow. The results arrive as a new test run in [Test Runs](https://docs.testdino.com/platform/playwright-test-runs), like any other test run from CI. > [!NOTE] > > Before you start, you need: > > * The **Pro**, **Team**, or **Enterprise** plan, and the Owner, Admin, or Member role. > * GitHub connected to the project. See [GitHub App](https://docs.testdino.com/integrations/ci-cd/github.md). > * Test cases linked to automated tests. See [Linked Automation Tests](https://docs.testdino.com/test-management/test-cases/list-view.md#linked-automation-tests). > * `@testdino/playwright` 2.11.1 or later and Playwright 1.56 or later in your project. > * A workflow set up once, as shown below. ## Set up your workflow Add a `testdino_selection` input to the GitHub Actions workflow that runs your Playwright tests, and pass it to `npx tdpw test`. Keep your existing triggers, setup, and install steps; the example shows only the lines to add. ```yaml .github/workflows/playwright.yml theme={null} on: workflow_dispatch: inputs: testdino_selection: { required: false } # set by TestDino's Run automation jobs: test: steps: - uses: actions/checkout@v4 - name: Run tests env: TESTDINO_TOKEN: ${{ secrets.TESTDINO_TOKEN }} TD_SELECTION: ${{ inputs.testdino_selection }} shell: bash run: | args=() if [ -n "$TD_SELECTION" ]; then args+=(--selection "$TD_SELECTION") fi npx tdpw test "${args[@]}" ``` * On a normal push the input is empty, so your whole suite runs as before. * Pass the input through `env:`, as in the example. * Add the workflow to every branch you want to run from TestDino. * Don't ignore the test command's result: no `|| true`, `|| echo`, `continue-on-error`, `| tee` without `pipefail`, or `set +e`. Otherwise the CI job shows green even when tests fail or nothing ran. * If the same workflow also has the [re-run inputs](https://docs.testdino.com/guides/debug-playwright-failures/configure-reruns-github-actions.md), set `testdino_rerun_from` to `required: false`. * Sharded workflows work. See [sharded runs](https://docs.testdino.com/cli/testdino-playwright-nodejs.md#sharded-runs) to report the shards as 1 test run. ## Start a run **1. Open Run automation** Click **Run automation** in 1 of 3 places. Each decides which test cases are ticked when the dialog opens. | Where | Ticked when it opens | | :- | :- | | Test Cases page, next to **+ New** | Nothing. Use **Select all** to run every linked test case. | | Test Cases selection bar, after you select test cases | The selected test cases that are linked | | The header of an open run (closed runs have no button) | The run's linked test cases that read `Not run in automation` | If nothing is linked yet, the button looks muted and tells you what to do when you click it. **2. Pick test cases** Tick the test cases to run. Only linked test cases can be ticked; others show `No linked automated test`. Use the suite checkboxes, **Select all**, **Clear**, and search to build the list, then click **Next**. A run takes up to 1,000 test cases. To change them later, click **Back** or the step names at the top. From a run, the chips **Not run in automation**, **Failed**, and **All automated** filter the list by each test case's CI result in that run. **3. Choose a branch and workflow** * **Branch**: the branch to run. From a run with a release branch, it starts on that branch; otherwise on the repository's default branch. Search the list or type a name. * **Workflow**: the workflows on that branch that are set up for TestDino. **Show why** explains the others and what to change. * **When a selected test is missing**: **Skip missing tests** (the default) or **Fail the CI job**. See [Tests missing on the branch](#tests-missing-on-the-branch). If some selected tests may not be on the branch, a `N may not run · View details` badge appears. Click it to see which ones and why. Those test cases are tagged `Test not found` or `Not found last run`; hover the tag for the reason. You can still run. If the workflow ignores the test command's result, a **Your workflow may hide test failures** box appears. Click **View fix** to see the file, the job, and the line to remove. The button shows how many tests will run, such as `Run 12 selected tests`. A test that runs in 3 browsers counts 3 times. **4. Run** Click the **Run** button. The dialog shows **Run started**. Click **View in GitHub** to open the workflow run, or **Done** to close the dialog. If GitHub is not connected yet, the dialog shows **Connect GitHub** instead of the branch and workflow. Connect it in the new tab, then click **Check again**. Your picks stay in the dialog. ## See the results The results arrive as a new test run in [Test Runs](https://docs.testdino.com/platform/playwright-test-runs) when the GitHub run finishes. It shows who started it, such as `Manual · `. When you started from a run, the run page also shows the progress: | The run page shows | Meaning | | :- | :- | | `Automated run started` | CI is running. On a run that reads the latest CI results, it clears once the results arrive. | | `Nothing ran on .` and a reason | CI ran nothing, for example because none of the selected tests are on that branch. | | `No results yet from the automated run started` | No results after 2 hours on a run that reads the latest CI results. Click **Check GitHub**. | If the run reads a specific CI run, the new results don't appear in it, and the message stays until you dismiss it. Click **Show latest results** to switch the run to the latest CI results. See [Pull automated results from CI](https://docs.testdino.com/test-management/manual-testing/manual-runs.md#pull-automated-results-from-ci). To retry only the tests that failed, open the new test run and click **Re-run**, then **This commit**. See [Re-run failed tests](https://docs.testdino.com/guides/debug-playwright-failures/rerun-failed-tests.md). ## Tests missing on the branch Before running, CI checks which selected tests exist on the branch. A test renamed, moved, or deleted on that branch is not found. | Situation | **Skip missing tests** | **Fail the CI job** | | :- | :- | :- | | Some tests are missing | Runs the rest. The CI job passes or fails based on those tests. | Runs the rest, then fails the CI job | | No selected test is found | Fails the CI job and runs nothing | Same | | The check itself cannot run | Runs the selection unchecked, with a warning in the CI log | Fails the CI job and runs nothing | The new test run shows a badge next to its title, such as `2 tests not found`. Hover it to see why, or click it to see the missing tests and the test cases that link them. A renamed or moved test needs to be linked again from its test case. ## Troubleshooting > **None of your workflows on \ can run selected tests yet.** > > No active workflow on the chosen branch has the `testdino_selection` input. Click **Show why** under **Workflow** for the reason, then [set up your workflow](#set-up-your-workflow) on that branch. > **TestDino isn't allowed to start workflows** > > The TestDino GitHub App needs **Actions** write permission. Ask an admin of your GitHub organization to approve it. See [GitHub App](https://docs.testdino.com/integrations/ci-cd/github.md). > **GitHub has no branch named \.** > > **Run** stays disabled when the branch is not on GitHub. Pick an existing branch, or push the branch first. > **None of the N selected tests were found, so nothing ran** > > None of the selected tests exist on the branch you ran: they were renamed, moved, or deleted there, or your Playwright arguments filter them out. Pick a branch that has the tests, or link the test cases again to the renamed tests. > **The CI job is green, but tests were missing or nothing ran** > > Your workflow ignores the exit code of `npx tdpw test`, for example with `|| true`, `|| echo`, or `continue-on-error`. Remove it so a failed check fails the CI job. The TestDino CLI log still lists the missing tests. Step 2 warns about this with **Your workflow may hide test failures**. > **error: unknown option '--selection'** > > The branch uses `@testdino/playwright` older than 2.11.1. Upgrade it on that branch. > **--selection needs Playwright 1.56 or newer** > > Upgrade `@playwright/test` to 1.56 or later on that branch. > **Selection \ is unknown or belongs to another project than this token's.** > > The workflow's `TESTDINO_TOKEN` is from a different TestDino project, or the selection id is wrong. Start the run from TestDino, and use an API key from the project you started it in. See [Generate API keys](https://docs.testdino.com/guides/generate-api-keys.md). > **This run may already have started** > > TestDino could not confirm the start with GitHub. Check the Actions tab in your repository before starting again, so the same tests do not run twice. ## Related > **[Link automated tests](https://docs.testdino.com/test-management/test-cases/list-view.md#linked-automation-tests)** > > Connect a test case to the automated test that covers it. > **[Runs](https://docs.testdino.com/test-management/manual-testing/manual-runs.md)** > > Track manual and automated test cases in 1 run. > **[Node.js CLI](https://docs.testdino.com/cli/testdino-playwright-nodejs.md)** > > Every `npx tdpw test` flag, including `--selection`. > **[Re-run failed tests](https://docs.testdino.com/guides/debug-playwright-failures/rerun-failed-tests.md)** > > Run only the failed or flaky test cases from a finished test run. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.