One action. One authority. Two boundary models.
The same deployment object is evaluated for authority conformance and substantive admissibility immediately before execution. The test reveals where the two systems complement each other and where they overlap.
Execution boundary
Authority can be restricted or invalidated. It cannot be expanded by runtime context.
Live proposed action
Receives the same authority envelope and proposed action. Paul maps the verifier's actual conformance result here.
Evaluates conformance result, tests, review evidence, policy, risk and current state.
Determines whether the exact action receives substantive clearance under all governing inputs.
Revalidates the bound action at the gateway and enforces the final decision.
Portable test object
This is the shared input. AbcadaCLV and VALO should not rewrite it to favour their own model.
Does AbcadaCLV's authority-conformance decision already cover every reason REHT uses to stop the action? If not, its result can become a necessary governed input to REHT. If yes, the two products substantially overlap at this boundary.