The platform is working correctly again. Runs will pick back up. Some runs may have failed that were in progress.

This was caused by a bad migration that didn't cause the deployment to fail automatically and so it rolled out to the instances.

The dashboard and API is down due to a deployment. We're working to fix this.