CBSs and networks in scope must be able to shut down in a controlled way (letting connected systems commit or roll back and leaving the whole integrated system safe, consistent and known), reset to their initial state, return to an earlier safe configuration, and restart from power-off by reloading fresh software and data from a read-only source in a time that suits their service without upsetting connected systems, with documentation aboard on how to do each after an incident.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.