Currently we don't have any CI job verifying that the agent-tui gets correctly displayed, in order to catch any eventual regression that may prevent the console to appear as expected (see https://issues.redhat.com/browse/OCPBUGS-11801).

A potential approach could be to deploy a CI jobs that interacts with the console, by testing a negative case (ie a wrong initial configuration) and fixing it, and then verifying that the installation completes as expected. For that, an inital PoC has been developed and it could be used as a reference: https://github.com/openshift-metal3/dev-scripts/pull/1516