The pure projection at the heart of runtime evaluation: given an
instance, its definition, the resolved actor/grants, and a snapshot,
compute "what can this actor do right now, and why not the rest." The
supplied snapshot is never refetched; guards using -> may call the
supplied EvaluateFromSnapshotArgs.guardDereference resolver.
Feed it a fresh snapshot (e.g. rebuilt from a live store on change) for
reactive re-evaluation. Best-effort by design: verdicts are advisory,
not enforcement.
The pure projection at the heart of runtime evaluation: given an instance, its definition, the resolved actor/grants, and a snapshot, compute "what can this actor do right now, and why not the rest." The supplied snapshot is never refetched; guards using
->may call the supplied EvaluateFromSnapshotArgs.guardDereference resolver. Feed it a fresh snapshot (e.g. rebuilt from a live store on change) for reactive re-evaluation. Best-effort by design: verdicts are advisory, not enforcement.