Full threadmunch-o-man·seems pretty interesting. So is this meant for actual container/workload orchestration or just monitoring/managing what is already running? Trying to figure out how this compares to something like Nomad.View on HN