CAPABILITIES

Eight control capabilities, organised by the problem they solve.

The public model is capability-first. Internal repositories and programmes appear only as provenance beneath the evidence.

Authority & Governance

Reasoning or intent must not silently become permission to cause a consequence.

Retries, races, stale context and changed material can turn one approved action into the wrong action or a duplicate effect.

Once consequential execution has started, a normal decision response is too late to serve as the interrupt mechanism.

After crash, partition or ambiguous acknowledgement, the external effect may be unknown and a blind retry may be unsafe.

Independent systems need enough shared context to cooperate without merging their authority domains.

A system can say that work completed while still hiding whether it was authorised, attempted, acknowledged, applied or reconciled.

Human approval can become stale or over-broad if it is not bound to the exact action and current state released for execution.

Software control claims can easily outrun what has actually been tested in simulators, real hardware or operational environments.