A.16.1:4.1 - Core shape
A conforming cue pack may publish:
cueNucleuspreservationRationale?laneCandidates?routeCandidateHints?valenceProfile?languageStateClosureDegreeRef?languageStateFacetProfileRef?detector?primaryAnchor?candidateAnchors?primaryWitnessRef?witnessRefs?exemplars?contrasts?traceRefs?embodimentRefs?modelStateRefs?scope?GammaTime?
cueNucleus names the minimal preserved core: what exactly is being kept visible rather than lost in carrier noise or premature endpoint wording.
primaryWitnessRef and primaryAnchor provide explicit triage when one witness or anchor is load-bearing for preservation. Secondary witnesses, anchors, traces, embodiment refs, and model-state refs may enrich the pack without displacing that primary nucleus.
laneCandidates and routeCandidateHints are early directional hints only. They are not a selected route, route rationale, or route-selection status. Those belong to RoutedCueSet under B.4.1.
The referenced facets keep their own definitions. primaryAnchor, candidateAnchors, contrasts, and exemplars commonly provide anchor material for AE under C.2.4; languageStateClosureDegreeRef docks to C.2.5; anchoring and representation-factor refs dock to C.2.6 and C.2.7; languageStateFacetProfileRef may bundle them through C.2.LS.
In this cluster, a cue is a salient epistemic nucleus extracted from witnesses, traces, felt tensions, model outputs, work-possibility hints, reliance-possibility hints, contrasts, or other grounds and made preservable as a pack. A raw signal-like trace counts as a cue only when that salience and preservability have been made explicit; otherwise it remains a trace, not yet a cue. Any evidence use is separately qualified under A.10 for a named target claim.