This documentation is available as Markdown for AI agents and LLMs. See the full Markdown index or append .md to any documentation URL.

Manage workflow runs in Expo Workflows

Edit page

Learn how to view, cancel, and re-run workflow runs in Expo Workflows.


Each time a workflow is triggered, Expo Workflows creates a workflow run. Use the Expo dashboard to follow a run and inspect its jobs. You can also cancel a run that is in progress, or create a new run from one that already finished.

Workflow run statuses

A workflow run has one of the following statuses:

  • New: Expo Workflows created the run, but it has not started yet.
  • Queued: The run is waiting for another run in its concurrency group. A queued run has no jobs until it starts.
  • In progress: One or more jobs are running.
  • Action required: A job is waiting for a manual action, such as an approval.
  • Success: All jobs completed successfully.
  • Failure: At least one job failed.
  • Canceled: The run was canceled before it completed.

Success, failure, and canceled are terminal statuses.

For the status values returned by the Expo Services REST API, see Get a workflow run.

View a workflow run

Open your project's workflows page and select a run. The run details show the status of each job, its logs and errors, and any outputs or artifacts that it produced.

You can also use EAS CLI to view a run. Run eas workflow:view to select a recent run, or provide a workflow run ID to view a specific run. Use eas workflow:logs to inspect logs. Use eas workflow:status WORKFLOW_RUN_ID --wait to follow a run until it completes.

To open a shell on the worker that runs a job, see Debug a workflow run with SSH.

Show the branch and commit of a run started from EAS CLI

By default, eas workflow:run uploads your local project files, so the run has no branch or commit. The dashboard shows it without a branch and does not list it under the branch filter. To show the branch and commit, push the commit to the GitHub repository linked to your project first. Then pass --git-branch and --git-commit-hash together:

Terminal
- eas workflow:run .eas/workflows/build.yml --git-branch "$(git branch --show-current)" --git-commit-hash "$(git rev-parse HEAD)"

The dashboard then links the run to that branch and commit. It also lists the run under the branch filter. An update job without branch or channel publishes to the Expo Update branch with the same name as the Git branch. The run still uses your uploaded project files, even if they differ from that commit. Expo Workflows does not post a commit status to GitHub. Workflow expressions see the branch and commit in the github context.

Cancel a workflow run

Open a run with a non-terminal status and select Cancel workflow. Jobs that already completed keep their final status. Expo Workflows stops the remaining work and marks the run as canceled.

To cancel a run with EAS CLI, run eas workflow:cancel WORKFLOW_RUN_ID. You can provide multiple workflow run IDs. If you do not provide an ID, select one or more in-progress or queued runs from the prompt.

Re-run a workflow

A re-run creates a new workflow run from the same workflow revision, project source, event data, and inputs as the original run. The dashboard links the related runs so that you can review their results. Re-running a workflow is only available on the Expo dashboard.

Re-run from failed jobs

Use this option to recover from a temporary failure without repeating successful jobs. For example, if a build job succeeds but its dependent submission job fails, the new run reuses the build and its build_id output, then retries the submission.

To re-run from failed jobs:

  1. Open the failed or canceled workflow run.
  2. Select Re-run workflow and then Re-run from failed jobs.
  3. Confirm the new run.

The new run handles jobs as follows:

  • Successful jobs are not run again. The new run reuses their results, outputs, logs, and artifacts.
  • Failed jobs run again.
  • Jobs that were blocked by a failed dependency are evaluated when that dependency completes. They run if their dependencies and conditions allow it.
  • Jobs skipped by an if condition remain skipped if the condition still evaluates to false.

You cannot select an arbitrary job as the starting point. Use Re-run from failed jobs to continue after failures, or use Re-run all jobs when you need every job to run again.

Re-run all jobs

Select Re-run workflow and then Re-run all jobs to create a new run that executes every job from the beginning. For a successful workflow, select Re-run workflow to do the same thing.

Use this option when the failure may depend on work completed by an earlier job, or when you need fresh outputs and artifacts from every job.

Re-run availability

A re-run can reuse source and artifacts only while Expo Services retains them. If the dashboard no longer shows the re-run options, start a new run with eas workflow:run .eas/workflows/my-workflow.yml.