It can be as simple as adding a line trigger to prevent a family (or suite) to continue any further as soon as a problem (abort) accurs:
family forecast trigger forecast != aborted
Alternatively a limit may play a similar role, as it may be reduced to 0 manually by an Analyst or an Operator, or even a dedicated batch script.
family forecast limit hpc 100 inlimit hpc # use: ecflow_client --alter change limit hpc 0 $path # to stop