The research question
A designated execution computer becomes unreachable, and an administrator appoints another device. The old computer may still be running workflows or regain network access later. If both retain valid external credentials, changing an internal authority version on the replacement does not physically prevent duplicate customer messages or conflicting source-system updates. Weakly connected-system research distinguishes tentative work from stabilized results. The proposed execution design adds business authorization and observable outcomes, because exchanging a record cannot by itself establish that an external action happened correctly.
A concrete design choice
The proposed manual handover first stops and isolates the former writer, revokes or changes its relevant external access, and reconciles outstanding commands and receipts. Only then is the new authority activated. If the old writer cannot be fenced, consequential execution remains paused rather than assuming it is dead because a timeout expired. A future automated design would need an external-action boundary that rejects stale authority, together with an appropriate coordination mechanism. The handover record should explain which actions remain uncertain so the replacement does not blindly replay them as fresh work.
What would demonstrate usefulness
Test a planned handover and a simulated network partition where the old executor remains alive. Verify that only the authorized active writer can produce external effects. Include pending commands with lost responses, because transferring ownership without reconciling them can create duplicates even when the two executors never overlap in time.