COMPO · 13.03 · TARGET DOMINANCE PROFILES

Observed is not desired.

13.03 формализует желаемую структуру доминанты отдельно от уже измеренной. Target profile задаётся явно: допустимые observed classifications, целевые лидеры, диапазоны leader/co-dominant/competing counts и при необходимости target margin range. Здесь ещё нет проверки удовлетворения цели и нет dominance score.

ModuleCompo 13.03
Targetexplicit
Observed13.02
Selectionexplicit only
Scorenone
NextSTOP

Target is an external desired state

00 / contract

13.02 describes what the current composition does. 13.03 describes what an explicit owner wants it to do. The current result cannot select its own target.

OBSERVED

Rank, margin and class from 13.02.

TARGET

Explicit desired structure.

BIND

Ensure same aggregation/policy context.

NO FIT YET

13.04 compares observed to target.

13.02 observed structure != desired structure explicit target profile + exact observed-policy identities → comparable target binding NOT YET: target satisfaction leader match score margin fit score correction synthesis final dominance score

Desired-structure descriptors only

01 / ownership
OWNS

Target profile

Desired class, leader identity semantics, count ranges and optional margin range.

OWNS

Feasibility + binding

Structural consistency and exact observed-policy context.

DEFERS

13.04

Observed-vs-target evaluation, scoring and correction synthesis.

No MetricResult vote.

owns_metrics=[]. Target ranges are desired descriptors, not scores.

Target profile kinds

02 / desired structure taxonomy
KindAllowed observed classification
SINGLE_DOMINANT_TARGETSINGLE_DOMINANT only
COMPETING_DOMINANTS_TARGETCOMPETING_DOMINANTS only
CO_DOMINANT_TARGETCO_DOMINANT only
FLEXIBLE_DOMINANCE_TARGETAny explicit non-empty subset of the three observed classes
No universal target.

COMPO does not assume that every composition should have one clear dominant. Single, competing and co-dominant structures are all valid target families when explicitly requested.

DominanceLeaderTarget

03 / who should lead

Target profiles can be structure-only or can explicitly name desired leading subjects.

DominanceLeaderTarget {
  mode:
    ANY_SUBJECT
    EXPLICIT_SUBJECT_SET,

  subjectRefs[],

  matchMode:
    ANY
    ANY_OF
    ALL_OF
    EXACT_SET
}
ANY_SUBJECT: subjectRefs = [] matchMode = ANY EXPLICIT_SUBJECT_SET: subjectRefs = non-empty canonical set matchMode ∈ ANY_OF | ALL_OF | EXACT_SET Target leader identity != focal candidate eligibility != observed rank mutation.

Desired count ranges

04 / structure bounds
DominanceCountRange {
  min,
  max
}

leaderCountRange:
  min >= 1

coDominantCandidateCountRange:
  min >= 1

competingCandidateCountRange:
  min >= 0

all:
  integer
  max >= min
  max <= 4096
Counts remain targets.

Leader count, co-dominant count and competing count are not numeric quality votes.

DominanceMarginTargetConstraint

05 / optional desired band
DominanceMarginTargetConstraint {
  mode:
    IGNORE
    REQUIRED_RANGE,

  min?,
  max?
}
IGNORE: min = null max = null REQUIRED_RANGE: 0 <= min <= max <= 1 IMPORTANT: target margin range does NOT redefine coDominantMaxMargin or singleDominantMinMargin from 13.02.

DominanceTargetProfile

06 / reusable explicit desired state
DominanceTargetProfile {
  targetProfileRef,

  profileKind,

  profileSourceRef,
  profileSourceOwnerRef,

  requiredAggregationProfileRef,
  requiredCompetitionPolicyRef,

  allowedObservedClassifications[],

  leaderTarget,

  leaderCountRange,
  coDominantCandidateCountRange,
  competingCandidateCountRange,

  marginTarget,

  semanticIntentRefs[],

  selectionMode: "EXPLICIT_ONLY",

  provenance,
  status,
  mutationAllowed: false
}
No self-selection.

An observed SINGLE_DOMINANT composition cannot automatically select SINGLE_DOMINANT_TARGET. Target selection requires an explicit source and owner.

Feasibility is not satisfaction

07 / structural validation

13.03 may reject an internally impossible target contract, but it does not evaluate whether the current composition meets that contract.

Feasibility checks: profile kind ↔ classification set leader target consistency EXACT_SET: leaderCountRange must contain exact target subject count ALL_OF: leaderCountRange.max must allow all target subjects coDominantCountRange.max must permit at least leaderCountRange.min required margin target must intersect at least one allowed classification interval under required competition policy feasible target != satisfied target

DominanceTargetProfileBinding

08 / same observed-policy context

A target is only comparable to observed data produced under the aggregation and competition policies it explicitly names.

DominanceTargetProfileBinding {
  targetProfileBindingRef,

  targetProfileRef,

  dominanceScopeRef,
  focalCandidateSetRef,
  descriptorSeedRef,

  attentionAttachmentRef,
  rankingAttachmentRef,

  focalRankingRef,
  dominanceMarginRef,
  competitionDescriptorRef,

  aggregationProfileRef,
  competitionPolicyRef,

  status,
  reasonCode,
  mutationAllowed: false
}
requiredAggregationProfileRef mismatch → N_A requiredCompetitionPolicyRef mismatch → N_A missing observed margin/classification → N_A malformed lineage → INVALID binding OK != target satisfied

Immutable target attachment

09 / descriptor chain
DominanceDescriptorTargetAttachment {
  targetAttachmentRef,

  rankingAttachmentRef,
  targetProfileBindingRef,
  targetProfileRef,

  status,
  mutationAllowed: false
}

13.03 appends a derived target layer. It does not mutate the 13.00 seed, 13.01 attention attachment or 13.02 ranking attachment.

Target selection is explicit only

10 / no circular objective
FORBIDDEN: observed SINGLE_DOMINANT → auto-select SINGLE_DOMINANT_TARGET observed CO_DOMINANT → auto-select CO_DOMINANT_TARGET rank 1 subject → infer desired leader candidate count → infer desired structure semantic role → infer target without explicit policy source + owner Observed structure cannot choose its own target.

Desired policy is identity-bearing

11 / deterministic lineage
targetProfileRef = hash( kind, source + owner, required aggregation profile, required competition policy, canonical allowed classes, leader target, three count ranges, margin target, canonical semantic refs, provenance ) targetProfileBindingRef = hash( target profile, scope / candidate set / seed, attention / ranking / margin / competition refs, observed aggregation + competition policy refs ) targetAttachmentRef = hash( ranking attachment, target profile binding, target profile )

Target does not rewrite observation

12 / firewalls
Forbidden

observed class → target inference; target leader → candidate injection; target profile → rank rewrite; target range → score; policy mismatch → zero.

Preserved

explicit selection; same-policy binding; feasible != satisfied; N_A != 0; descriptor != score; module order != priority.

Topic13 ownership chain

13 / integration
OwnerCanonical role
13.00scope, role evidence/resolution, focal candidate set
13.01attention factor matrix
13.02explicit aggregation, rank, observed margin and competition class
13.03desired target dominance profiles + comparable binding
13.04observed-vs-target scoring, correction and final tests

Wrong target policy → owner-qualified correction

14 / repair
DominanceTargetProfileCorrectionRequest {
  requestRef,
  targetOwnerRef,

  reasonCode,
  evidenceRefs[],
  suggestedAction,

  status,
  mutationAllowed: false
}

CANON → LAB → DATA → IDEAS

15 / mandatory hooks

Target profiles remain explicit policy objects. Human studies may reveal better profile bands or leader expectations, but LAB results do not auto-mutate CANON.

IDEAS_REFS[]7 R&D refs retained.
DATA_CAPTURE[]target kind, leader, ranges, policy refs and binding.
DEFERRED_SEMANTIC_INPUTS[]semantic desirability stays explicit/deferred.
HUMAN_FEEDBACK_HOOKS[]target-kind/leader/range/policy corrections.
CONFLICT_PRIORITY_ASSUMPTIONS[]target profile never becomes hidden priority.
LAB_HOOKS[]single/competing/co-dominant target studies.

Human correction reasons

16 / evidence
FamilyExamples
ProfileTARGET_PROFILE_KIND_WRONG, TARGET_CLASSIFICATION_WRONG
LeaderTARGET_LEADER_WRONG, TARGET_LEADER_MATCH_MODE_WRONG
CountsTARGET_LEADER/CO_DOMINANT/COMPETING_COUNT_TOO_LOW/HIGH
MarginTARGET_MARGIN_TOO_LOW/HIGH
ContextTARGET_AGGREGATION_PROFILE_WRONG, TARGET_COMPETITION_POLICY_WRONG
StrictnessTARGET_PROFILE_SHOULD_BE_FLEXIBLE / STRICTER

Visual laboratory hook

17 / observed vs desired

The same observed ranking can be judged against different explicit target profiles. 13.03 stores those targets; 13.04 will evaluate fit.

ONE OBSERVED STRUCTURE · MULTIPLE POSSIBLE EXPLICIT TARGETS OBSERVED classification: COMPETING_DOMINANTS TARGET A SINGLE_DOMINANT_TARGET leader: explicit subject A TARGET B COMPETING_DOMINANTS_TARGET margin target: explicit range 13.03 stores target contracts · it does NOT select a target from observed data observed ↔ target evaluation and scoring → 13.04

13.03 regression suite

18 / deterministic tests

Tests cover profile-kind/classification consistency, explicit leader semantics, count/margin ranges, policy-dependent feasibility, same-policy binding, explicit-only selection, immutable attachment and no target-satisfaction leakage into 13.03.

Tests not run yet.

DATA FOR LAYOUT ENGINE

19 / machine contract

Machine-readable 13.03 canonical contract.

{
  "module": "Compo 13.03",
  "name": "target_dominance_profiles",
  "version": "1.0-canonical",
  "canonical_main_title": "Target dominance profiles",
  "topic": "13_ДОМИНАНТА_И_ГЛАВНЫЙ_ФОКУС",
  "scope": "EXPLICIT_DESIRED_DOMINANCE_STRUCTURE_PROFILES_FEASIBILITY_BINDING_AND_IMMUTABLE_TARGET_ATTACHMENT_ONLY",
  "main_topic_contract": {
    "topic_title": "Доминанта и главный фокус",
    "topic_goal": "Define dominance as a measurable leader of attention through focal candidates, dominance margin and competing/co-dominant cases without assuming a universal mandatory single focal center.",
    "expected_output": "DominanceDescriptor, focal rank, dominance margin, target profiles",
    "planned_files": [
      "13.00 Dominant role / focal candidate model",
      "13.01 Attention factors: size, contrast, isolation, position",
      "13.02 Dominance margin и competing dominants",
      "13.03 Target dominance profiles",
      "13.04 Scoring, correction и tests"
    ]
  },
  "ownership": {
    "owner": "Compo13.03",
    "owns": [
      "DominanceTargetProfileKind taxonomy",
      "DominanceLeaderTarget",
      "DominanceCountRange",
      "DominanceMarginTargetConstraint",
      "DominanceTargetProfile",
      "DominanceTargetProfileFeasibility",
      "DominanceTargetProfileBinding",
      "DominanceDescriptorTargetAttachment",
      "explicit target-profile selection contract",
      "non-mutating target-profile correction request"
    ],
    "does_not_own": [
      "DominanceScope / focal candidate eligibility — Compo13.00",
      "attention factor measurements/response contexts — Compo13.01",
      "attention aggregation / focal rank / observed dominance margin — Compo13.02",
      "observed competition classification — Compo13.02",
      "target-profile satisfaction evaluation — Compo13.04",
      "target-profile scoring — Compo13.04",
      "dominance correction synthesis — Compo13.04",
      "final independent Topic13 vote — Compo13.04 if canonically justified",
      "global hierarchy semantics",
      "global rule/conflict priority",
      "foreign geometry or layout mutation"
    ]
  },
  "consumes": [
    "valid Compo13.00 DominanceScope",
    "valid Compo13.00 FocalCandidateSet",
    "valid Compo13.00 DominanceDescriptorSeed",
    "valid Compo13.01 DominanceDescriptorAttentionAttachment",
    "valid Compo13.02 AttentionAggregationProfile identity",
    "valid Compo13.02 FocalRanking",
    "valid Compo13.02 DominanceMarginDescriptor",
    "valid Compo13.02 DominanceCompetitionPolicy",
    "valid Compo13.02 DominanceCompetitionDescriptor",
    "valid Compo13.02 DominanceDescriptorRankingAttachment",
    "explicit DominanceTargetProfile"
  ],
  "provides": [
    "DominanceLeaderTarget",
    "DominanceCountRange",
    "DominanceMarginTargetConstraint",
    "DominanceTargetProfile",
    "DominanceTargetProfileFeasibility",
    "DominanceTargetProfileBinding",
    "DominanceDescriptorTargetAttachment",
    "Topic13.03 observation record"
  ],
  "owns_metrics": [],
  "metric_ownership_note": "13.03 owns desired-structure descriptors only. It does not compare observed versus target, does not produce MetricResult, and owns no independent numeric vote.",
  "target_profile_kind_taxonomy": {
    "SINGLE_DOMINANT_TARGET": "desired observed classification is SINGLE_DOMINANT",
    "COMPETING_DOMINANTS_TARGET": "desired observed classification is COMPETING_DOMINANTS",
    "CO_DOMINANT_TARGET": "desired observed classification is CO_DOMINANT",
    "FLEXIBLE_DOMINANCE_TARGET": "explicit non-empty subset of observed classifications is acceptable"
  },
  "leader_target_schema": {
    "mode": "ANY_SUBJECT|EXPLICIT_SUBJECT_SET",
    "subjectRefs": "empty for ANY_SUBJECT; non-empty canonical set for EXPLICIT_SUBJECT_SET",
    "matchMode": "ANY for ANY_SUBJECT; ANY_OF|ALL_OF|EXACT_SET for EXPLICIT_SUBJECT_SET",
    "firewall": "leader target declares desired identity only; it does not alter focal candidate eligibility or observed rank"
  },
  "count_range_schema": {
    "min": "integer >=0",
    "max": "integer >= min and <=4096",
    "leaderCountRange": "min >=1",
    "coDominantCandidateCountRange": "min >=1 because 13.02 coDominantCandidateRefs includes exact leaders",
    "competingCandidateCountRange": "min >=0 because 13.02 competingCandidateRefs excludes exact leaders",
    "firewall": "count target is a desired range, not a score"
  },
  "margin_target_constraint_schema": {
    "mode": "IGNORE|REQUIRED_RANGE",
    "min": "null for IGNORE; finite 0..1 for REQUIRED_RANGE",
    "max": "null for IGNORE; finite min..1 for REQUIRED_RANGE",
    "firewall": "margin target does not redefine 13.02 competition thresholds"
  },
  "target_profile_schema": {
    "targetProfileRef": "stable identity",
    "profileKind": "closed target-profile taxonomy",
    "profileSourceRef": "explicit source",
    "profileSourceOwnerRef": "owner",
    "requiredAggregationProfileRef": "exact 13.02 aggregation profile identity",
    "requiredCompetitionPolicyRef": "exact 13.02 competition interpretation policy identity",
    "allowedObservedClassifications": "explicit non-empty classification set",
    "leaderTarget": "DominanceLeaderTarget",
    "leaderCountRange": "DominanceCountRange",
    "coDominantCandidateCountRange": "DominanceCountRange",
    "competingCandidateCountRange": "DominanceCountRange",
    "marginTarget": "DominanceMarginTargetConstraint",
    "semanticIntentRefs": "optional opaque refs",
    "provenance": "canonical provenance class",
    "status": "OK|INVALID",
    "mutationAllowed": false,
    "selectionMode": "EXPLICIT_ONLY"
  },
  "kind_classification_contract": {
    "SINGLE_DOMINANT_TARGET": [
      "SINGLE_DOMINANT"
    ],
    "COMPETING_DOMINANTS_TARGET": [
      "COMPETING_DOMINANTS"
    ],
    "CO_DOMINANT_TARGET": [
      "CO_DOMINANT"
    ],
    "FLEXIBLE_DOMINANCE_TARGET": "any explicit non-empty subset of CO_DOMINANT|COMPETING_DOMINANTS|SINGLE_DOMINANT"
  },
  "profile_feasibility": {
    "purpose": "structural consistency only; it does not evaluate current observed layout against the target",
    "checks": [
      "profile kind/classification contract",
      "leader target internal consistency",
      "leader count compatibility with EXACT_SET/ALL_OF",
      "co-dominant count range can contain at least the minimum leader count",
      "required margin range intersects at least one allowed classification interval under requiredCompetitionPolicyRef",
      "all ranges are finite/integer and bounded"
    ],
    "status": "OK|INVALID",
    "firewall": "feasible target != satisfied target"
  },
  "classification_intervals_under_policy": {
    "CO_DOMINANT": "[0, coDominantMaxMargin]",
    "COMPETING_DOMINANTS": "(coDominantMaxMargin, singleDominantMinMargin)",
    "SINGLE_DOMINANT": "[singleDominantMinMargin, 1]",
    "note": "intervals are interpretation-policy dependent; 13.03 does not create or modify those thresholds"
  },
  "target_profile_binding_schema": {
    "targetProfileBindingRef": "stable identity",
    "targetProfileRef": "profile",
    "dominanceScopeRef": "13.00 scope",
    "focalCandidateSetRef": "13.00 candidate set",
    "descriptorSeedRef": "13.00 seed",
    "attentionAttachmentRef": "13.01 attachment",
    "rankingAttachmentRef": "13.02 attachment",
    "focalRankingRef": "13.02 ranking",
    "dominanceMarginRef": "13.02 margin",
    "competitionDescriptorRef": "13.02 observed class",
    "aggregationProfileRef": "must equal profile.requiredAggregationProfileRef",
    "competitionPolicyRef": "must equal profile.requiredCompetitionPolicyRef",
    "status": "OK|N_A|INVALID",
    "reasonCode": "explicit",
    "mutationAllowed": false,
    "N_A": [
      "required aggregation profile does not match observed ranking profile",
      "required competition policy does not match observed classification policy",
      "required margin/classification artifact is unavailable"
    ],
    "firewall": "binding means comparable context exists; it does not mean target satisfied"
  },
  "target_attachment_schema": {
    "targetAttachmentRef": "stable identity",
    "rankingAttachmentRef": "13.02 attachment",
    "targetProfileBindingRef": "13.03 binding",
    "targetProfileRef": "profile",
    "status": "OK|N_A|INVALID",
    "mutationAllowed": false,
    "firewall": "13.03 creates a derived target attachment and never mutates 13.00/13.01/13.02 artifacts"
  },
  "selection_policy": {
    "mode": "EXPLICIT_ONLY",
    "noAutomaticSelection": true,
    "forbiddenInferences": [
      "observed SINGLE_DOMINANT => select SINGLE_DOMINANT_TARGET",
      "observed CO_DOMINANT => select CO_DOMINANT_TARGET",
      "rank 1 subject => infer desired leader",
      "semantic role => infer target profile without explicit policy owner",
      "candidate count => infer desired structure"
    ],
    "firewall": "observed structure cannot choose its own target"
  },
  "identity_policy": {
    "targetProfileRef": "hash(kind, source/owner, required aggregation/competition policy refs, canonical classifications, leader target, count ranges, margin target, canonical semantic refs, provenance)",
    "targetProfileBindingRef": "hash(profile + scope/set/seed/attention/ranking/margin/competition refs + observed aggregation/policy refs)",
    "targetAttachmentRef": "hash(rankingAttachmentRef + targetProfileBindingRef + targetProfileRef)",
    "set_semantics": "classification/subject/semantic arrays canonicalized",
    "excluded_from_identity": [
      "display label",
      "display color",
      "free-form comment",
      "confidence"
    ]
  },
  "status_policy": {
    "targetProfile": "OK|INVALID",
    "profileFeasibility": "OK|INVALID",
    "targetProfileBinding": "OK|N_A|INVALID",
    "targetAttachment": "OK|N_A|INVALID",
    "profileMismatch": "N_A, not zero",
    "missingObservedMarginOrClass": "N_A, not failed score",
    "N_A_not_zero": true
  },
  "critical_firewalls": [
    "observed dominance structure != desired dominance structure",
    "observed classification != target profile",
    "target profile selection must be explicit",
    "observed SINGLE_DOMINANT does not imply SINGLE_DOMINANT_TARGET",
    "observed COMPETING_DOMINANTS does not imply COMPETING_DOMINANTS_TARGET",
    "observed CO_DOMINANT does not imply CO_DOMINANT_TARGET",
    "rank 1 subject != desired leader",
    "target leader identity does not alter focal candidate eligibility",
    "target leader identity does not alter observed focal rank",
    "target profile != score",
    "feasible target profile != satisfied target profile",
    "target profile binding != target satisfaction",
    "leader count target != score",
    "co-dominant count target != score",
    "competing count target != score",
    "margin target != score",
    "margin target does not redefine competition thresholds",
    "requiredAggregationProfileRef must be explicit",
    "requiredCompetitionPolicyRef must be explicit",
    "aggregation-profile mismatch = N_A",
    "competition-policy mismatch = N_A",
    "N_A != 0",
    "missing observed margin != target failure score",
    "no universal single-focus target",
    "no universal co-dominant target",
    "no universal competing-dominants target",
    "no universal margin target",
    "no universal leader count target",
    "semanticIntentRefs remain opaque",
    "profileKind does not create global hierarchy",
    "target profile is local to its explicit observed-policy identities",
    "profile satisfaction evaluation is deferred to 13.04",
    "dominance scoring is deferred to 13.04",
    "correction synthesis is deferred to 13.04",
    "13.03 owns no independent numeric vote",
    "target attachment != final DominanceDescriptor",
    "local != global",
    "descriptor != score",
    "module execution order != rule priority",
    "correction request != foreign mutation",
    "no universal beautyScore",
    "IDEA != CANON",
    "EXPERIMENT != CANON"
  ],
  "correction_request_schema": {
    "requestRef": "stable identity",
    "targetOwnerRef": "owner whose target profile/policy should be reviewed",
    "reasonCode": "explicit",
    "evidenceRefs": "canonicalized refs",
    "suggestedAction": "non-mutating suggestion",
    "status": "OPEN|INVALID",
    "mutationAllowed": false
  },
  "deferred_downstream": {
    "13.04": [
      "observed-versus-target condition evaluation",
      "leader identity match evaluation",
      "classification target satisfaction",
      "leader/co-dominant/competing count fit",
      "margin-range fit",
      "profile satisfaction composite",
      "dominance correction synthesis",
      "independent Topic13 vote if canonically justified",
      "consolidated Topic13 tests"
    ]
  },
  "IDEAS_REFS": [
    "IDEA-01 full human test after all 72 topics; test which desired dominance structures humans perceive as intentional",
    "IDEA-02 semantics needed before placement; target semanticIntentRefs remain opaque and explicit",
    "IDEA-03 explicit rule priorities; target profile identity must not become hidden global priority",
    "IDEA-04 collect structured data for future composition model",
    "IDEA-05 preserve reasons for human corrections",
    "IDEA-06 structured machine-readable layer for website",
    "IDEA-07 graphics-for-post-text research remains downstream/separate"
  ],
  "DATA_CAPTURE": [
    "sceneRef",
    "targetProfileRef",
    "profileKind",
    "profileSourceRef",
    "requiredAggregationProfileRef",
    "requiredCompetitionPolicyRef",
    "allowedObservedClassifications[]",
    "leaderTarget.mode",
    "leaderTarget.subjectRefs[]",
    "leaderTarget.matchMode",
    "leaderCountRange",
    "coDominantCandidateCountRange",
    "competingCandidateCountRange",
    "marginTarget",
    "semanticIntentRefs[]",
    "targetProfileBindingRef",
    "bindingStatus",
    "bindingReasonCode",
    "targetAttachmentRef",
    "humanReasonCodes[]",
    "future dominanceEvaluationRef"
  ],
  "DEFERRED_SEMANTIC_INPUTS": [
    "semantic desirability of one versus multiple focal centers",
    "which content role should lead",
    "brand/editorial desired hierarchy",
    "role-specific target margin meaning",
    "future global conflict-priority refs"
  ],
  "HUMAN_FEEDBACK_HOOKS": [
    "TARGET_PROFILE_KIND_WRONG",
    "TARGET_CLASSIFICATION_WRONG",
    "TARGET_LEADER_WRONG",
    "TARGET_LEADER_MATCH_MODE_WRONG",
    "TARGET_LEADER_COUNT_TOO_LOW",
    "TARGET_LEADER_COUNT_TOO_HIGH",
    "TARGET_CO_DOMINANT_COUNT_TOO_LOW",
    "TARGET_CO_DOMINANT_COUNT_TOO_HIGH",
    "TARGET_COMPETING_COUNT_TOO_LOW",
    "TARGET_COMPETING_COUNT_TOO_HIGH",
    "TARGET_MARGIN_TOO_LOW",
    "TARGET_MARGIN_TOO_HIGH",
    "TARGET_AGGREGATION_PROFILE_WRONG",
    "TARGET_COMPETITION_POLICY_WRONG",
    "TARGET_PROFILE_SHOULD_BE_FLEXIBLE",
    "TARGET_PROFILE_SHOULD_BE_STRICTER"
  ],
  "CONFLICT_PRIORITY_ASSUMPTIONS": [
    "module execution order is not priority",
    "target profile selection is explicit and not inferred from observed state",
    "profileKind is a desired-structure label, not global rule priority",
    "leader target subject order is non-semantic",
    "classification list order is non-semantic",
    "future reviewed global conflict-priority owner remains authoritative"
  ],
  "LAB_HOOKS": [
    "human A/B: same observed ranking under different explicit target profiles",
    "human study: desired single focus versus acceptable competing focus",
    "human study: desired co-dominant pairs/groups",
    "target margin calibration without auto-writing universal bands into canon",
    "leader-identity studies: expected semantic leader versus observed visual leader",
    "strict versus flexible target-profile comparison"
  ],
  "observation_record": {
    "moduleRef": "Compo13.03",
    "metricResults": [],
    "dominanceEvaluationRefs": []
  },
  "stop_condition": "STOP after 13.03. Do not create 13.04 in this module."
}
VISITS / ...