@tangle-network/agent-eval 0.131.1 → 0.133.0

package/dist/belief-state/index.d.ts

@@ -1,622 +0,0 @@
-import { a as RunRecord, s as RunSplitTag } from "../run-record-BJnYdTxO.js";
-import { s as TraceStore } from "../store-CT9YIIve.js";
-import { w as CalibrationReport } from "../index-D_F6VAKe.js";
-import { i as OffPolicyTrajectory, n as OffPolicyEstimate, r as OffPolicyOptions } from "../off-policy-mskQw8Mb.js";
-import { i as CodeAgentSessionMetrics, n as CodeAgentSessionIntakeOptions, t as CodeAgentSessionDiagnostic, v as CodeAgentSessionObservation, y as CodeAgentSessionSource } from "../code-agent-session-aYa3SGKz.js";
-import { a as RuntimeTrajectoryRecord, n as RuntimeTrajectoryEvidenceProjection, t as ProjectRuntimeTrajectoryEvidenceOptions } from "../runtime-trajectory-BXxG4lyi.js";
-//#region src/belief-state/types.d.ts
-declare const BELIEF_DECISION_KINDS: readonly ['continue', 'verify', 'ask', 'retry', 'stop', 'memory-write', 'memory-read', 'tool-select', 'skill-select', 'workflow-select', 'surface-promote'];
-type BeliefDecisionKind = (typeof BELIEF_DECISION_KINDS)[number];
-declare const BELIEF_EVIDENCE_SOURCES: readonly ['run', 'span', 'event', 'finding', 'memory', 'knowledge', 'policy'];
-type BeliefEvidenceSource = (typeof BELIEF_EVIDENCE_SOURCES)[number];
-declare const BELIEF_EVIDENCE_QUALITIES: readonly ['direct', 'derived', 'self-reported', 'unverified', 'stale', 'contradicted'];
-type BeliefEvidenceQuality = (typeof BELIEF_EVIDENCE_QUALITIES)[number];
-declare const BELIEF_EVALUATION_CRITERIA: readonly [{
-  readonly id: 'capture-integrity';
-  readonly label: 'Capture integrity';
-  readonly reasonCodes: readonly ['trace-missing', 'run-record-missing', 'backend-integrity-missing'];
-}, {
-  readonly id: 'decision-completeness';
-  readonly label: 'Decision completeness';
-  readonly reasonCodes: readonly ['candidate-actions-missing', 'chosen-action-missing', 'decision-evidence-missing'];
-}, {
-  readonly id: 'evidence-quality';
-  readonly label: 'Evidence quality';
-  readonly reasonCodes: readonly ['evidence-stale', 'evidence-contradictory', 'evidence-unverified', 'evidence-self-reported'];
-}, {
-  readonly id: 'outcome-quality';
-  readonly label: 'Outcome quality';
-  readonly reasonCodes: readonly ['outcome-missing', 'outcome-delayed', 'cost-missing'];
-}, {
-  readonly id: 'calibration';
-  readonly label: 'Calibration';
-  readonly reasonCodes: readonly ['confidence-missing', 'calibration-unsupported', 'calibration-gap-high'];
-}, {
-  readonly id: 'accepted-region-risk';
-  readonly label: 'Accepted-region risk';
-  readonly reasonCodes: readonly ['accepted-error-high', 'coverage-too-low'];
-}, {
-  readonly id: 'policy-value';
-  readonly label: 'Policy value';
-  readonly reasonCodes: readonly ['utility-lift-missing', 'baseline-dominates', 'cost-too-high'];
-}, {
-  readonly id: 'ope-support';
-  readonly label: 'OPE support';
-  readonly reasonCodes: readonly ['behavior-propensity-missing', 'behavior-propensity-invalid', 'target-propensity-missing', 'target-propensity-invalid', 'effective-sample-size-low', 'importance-weight-high'];
-}, {
-  readonly id: 'memory-health';
-  readonly label: 'Memory health';
-  readonly reasonCodes: readonly ['memory-stale', 'memory-poisoning-risk', 'context-bloat', 'memory-write-unverified'];
-}, {
-  readonly id: 'surface-attribution';
-  readonly label: 'Surface attribution';
-  readonly reasonCodes: readonly ['surface-claim-unsupported', 'causal-attribution-missing'];
-}, {
-  readonly id: 'generalization';
-  readonly label: 'Generalization';
-  readonly reasonCodes: readonly ['split-missing', 'holdout-regression', 'task-family-coverage-low', 'leakage-risk'];
-}, {
-  readonly id: 'promotion';
-  readonly label: 'Promotion';
-  readonly reasonCodes: readonly ['negative-control-failed', 'promotion-gate-failed', 'human-review-required'];
-}];
-type BeliefEvaluationCriterionId = (typeof BELIEF_EVALUATION_CRITERIA)[number]['id'];
-type BeliefDecisionReasonCode = (typeof BELIEF_EVALUATION_CRITERIA)[number]['reasonCodes'][number];
-interface BeliefDecisionReason {
-  code: BeliefDecisionReasonCode;
-  criterion?: BeliefEvaluationCriterionId;
-  detail?: string;
-  evidenceIds?: string[];
-  metadata?: Record<string, unknown>;
-}
-declare function isBeliefDecisionKind(value: unknown): value is BeliefDecisionKind;
-declare function isBeliefEvidenceSource(value: unknown): value is BeliefEvidenceSource;
-interface BeliefEvidenceRef {
-  source: BeliefEvidenceSource;
-  id: string;
-  runId?: string;
-  spanId?: string;
-  eventId?: string;
-  detail?: string;
-  quality?: BeliefEvidenceQuality;
-  observedAt?: string;
-  metadata?: Record<string, unknown>;
-}
-interface BeliefDecisionOutcome {
-  success?: boolean;
-  score?: number;
-  reward?: number;
-  costUsd?: number;
-  observedAt?: string;
-  metadata?: Record<string, unknown>;
-}
-interface BeliefDecisionPoint {
-  id: string;
-  runId: string;
-  scenarioId?: string;
-  stepIndex: number;
-  kind: BeliefDecisionKind;
-  chosenAction: string;
-  candidateActions?: string[];
-  confidence?: number;
-  behaviorProb?: number;
-  targetProb?: number;
-  qHatChosen?: number | null;
-  vHatTarget?: number | null;
-  costUsd?: number;
-  evidence: BeliefEvidenceRef[];
-  outcome?: BeliefDecisionOutcome;
-  reasons?: BeliefDecisionReason[];
-  metadata?: Record<string, unknown>;
-}
-interface BeliefDecisionExtractionDiagnostic {
-  runId: string;
-  eventId?: string;
-  severity: 'info' | 'warning' | 'error';
-  reason: string;
-}
-interface BeliefDecisionExtractionReport {
-  decisions: BeliefDecisionPoint[];
-  diagnostics: BeliefDecisionExtractionDiagnostic[];
-}
-type BeliefPolicyAction = 'accept' | 'defer' | 'verify' | 'ask' | 'retry' | 'stop';
-interface BeliefPolicyDecision {
-  action: BeliefPolicyAction;
-  confidence?: number;
-  targetProb?: number;
-  qHatChosen?: number | null;
-  vHatTarget?: number | null;
-  reason?: string;
-  reasons?: BeliefDecisionReason[];
-}
-interface BeliefSelectivePolicy {
-  id: string;
-  decide(point: BeliefDecisionPoint): BeliefPolicyDecision;
-}
-interface BeliefOpeTargetPolicy {
-  id: string;
-  targetProbOf(point: BeliefDecisionPoint): number | null | undefined;
-  qHatChosenOf?(point: BeliefDecisionPoint): number | null | undefined;
-  vHatTargetOf?(point: BeliefDecisionPoint): number | null | undefined;
-}
-interface BeliefUtilityOptions {
-  successUtility?: number;
-  failureUtility?: number;
-  deferUtility?: number;
-  verifyCost?: number;
-  askCost?: number;
-  retryCost?: number;
-  stopUtility?: number;
-  costWeight?: number;
-}
-interface BeliefSelectivePolicyMetrics {
-  policyId: string;
-  n: number;
-  accepted: number;
-  rejected: number;
-  coverage: number;
-  acceptedErrorRate: number;
-  baselineUtility: number;
-  policyUtility: number;
-  utilityDelta: number;
-  utilityCi95: {
-    mean: number;
-    lower: number;
-    upper: number;
-  };
-  rejectedMeanReward: number | null;
-  recommendation: 'ship' | 'hold' | 'need_more_data';
-  reasons: string[];
-}
-interface BeliefOpeSupportDiagnostics {
-  supported: boolean;
-  n: number;
-  dropped: number;
-  effectiveSampleSize: number;
-  effectiveSampleRatio: number;
-  maxImportanceWeight: number;
-  reasons: string[];
-}
-interface BeliefOpeReport {
-  targetPolicyId: string;
-  ips: OffPolicyEstimate;
-  snips: OffPolicyEstimate;
-  dr: OffPolicyEstimate;
-  support: BeliefOpeSupportDiagnostics;
-}
-type BeliefEvaluationStatus = 'ship' | 'hold' | 'need_more_data';
-type BeliefCalibrationStatus = 'supported' | 'unsupported';
-type BeliefOpeStatus = 'supported' | 'unsupported' | 'not_requested';
-interface BeliefPolicyEvaluationReport {
-  policyId: string;
-  n: number;
-  status: BeliefEvaluationStatus;
-  selectiveStatus: BeliefEvaluationStatus;
-  calibrationStatus: BeliefCalibrationStatus;
-  opeStatus: BeliefOpeStatus;
-  opeTargetPolicyId?: string;
-  selective: BeliefSelectivePolicyMetrics;
-  calibration?: CalibrationReport;
-  ope?: BeliefOpeReport;
-  diagnostics: string[];
-}
-//#endregion
-//#region src/belief-state/calibration.d.ts
-type BeliefCalibrationRegion = 'all' | 'accepted' | 'rejected';
-interface BeliefCalibrationOptions {
-  bins?: number;
-  minPairs?: number;
-  policy?: BeliefSelectivePolicy;
-  region?: BeliefCalibrationRegion;
-}
-declare function calibrateBeliefDecisions(points: BeliefDecisionPoint[], options?: BeliefCalibrationOptions): CalibrationReport | null;
-//#endregion
-//#region src/belief-state/ope.d.ts
-interface BeliefOpeOptions extends OffPolicyOptions {
-  minEffectiveSampleSize?: number;
-  minEffectiveSampleRatio?: number;
-  maxDiagnostics?: number;
-}
-interface BeliefOffPolicyTrajectoryReport {
-  targetPolicyId: string;
-  trajectories: OffPolicyTrajectory[];
-  dropped: number;
-  diagnostics: string[];
-}
-declare function embeddedBeliefOpeTargetPolicy(id?: string): BeliefOpeTargetPolicy;
-declare function beliefDecisionsToOffPolicyTrajectories(points: BeliefDecisionPoint[], targetPolicy: BeliefOpeTargetPolicy, options?: Pick<BeliefOpeOptions, 'maxDiagnostics'>): BeliefOffPolicyTrajectoryReport;
-declare function evaluateBeliefOffPolicy(points: BeliefDecisionPoint[], targetPolicy: BeliefOpeTargetPolicy, options?: BeliefOpeOptions): BeliefOpeReport;
-//#endregion
-//#region src/belief-state/selective.d.ts
-interface EvaluateBeliefSelectivePolicyOptions {
-  utility?: BeliefUtilityOptions;
-  minN?: number;
-  minAccepted?: number;
-  minUtilityDelta?: number;
-  seed?: number;
-}
-declare function thresholdSelectivePolicy(options: {
-  id?: string;
-  confidenceThreshold: number;
-  belowThresholdAction?: Exclude<BeliefPolicyAction, 'accept'>;
-}): BeliefSelectivePolicy;
-declare function evaluateBeliefSelectivePolicy(points: BeliefDecisionPoint[], policy: BeliefSelectivePolicy, options?: EvaluateBeliefSelectivePolicyOptions): BeliefSelectivePolicyMetrics;
-//#endregion
-//#region src/belief-state/report.d.ts
-interface AnalyzeBeliefPolicyOpeOptions extends BeliefOpeOptions {
-  targetPolicy?: BeliefOpeTargetPolicy;
-}
-interface AnalyzeBeliefPolicyOptions {
-  points: BeliefDecisionPoint[];
-  policy: BeliefSelectivePolicy;
-  selective?: EvaluateBeliefSelectivePolicyOptions;
-  calibration?: BeliefCalibrationOptions;
-  ope?: AnalyzeBeliefPolicyOpeOptions;
-  requireOpe?: boolean;
-}
-declare function analyzeBeliefPolicy(options: AnalyzeBeliefPolicyOptions): BeliefPolicyEvaluationReport;
-//#endregion
-//#region src/belief-state/code-agent-corpus.d.ts
-type CodeAgentBeliefDecisionTargetId = 'failure-recovery' | 'tool-selection' | 'graph-completion';
-interface ExtractCodeAgentBeliefDecisionPointsOptions {
-  source: CodeAgentSessionSource;
-  entries: unknown[];
-  /** Reuse intake's provider-neutral projection when available. */
-  observation?: CodeAgentSessionObservation;
-  run: Pick<RunRecord, 'runId' | 'scenarioId' | 'outcome' | 'costUsd'>;
-  sourcePath?: string;
-}
-interface BeliefDecisionInventoryBucket {
-  id: string;
-  kind?: BeliefDecisionKind;
-  targetId?: CodeAgentBeliefDecisionTargetId;
-  n: number;
-  withOutcome: number;
-  withConfidence: number;
-  withCandidateActions: number;
-  withBehaviorProb: number;
-  withTargetProb: number;
-  successRate: number | null;
-  meanScore: number | null;
-  meanConfidence: number | null;
-}
-interface BeliefDecisionInventoryReport {
-  n: number;
-  byKind: BeliefDecisionInventoryBucket[];
-  byTarget: BeliefDecisionInventoryBucket[];
-  diagnostics: string[];
-}
-interface BeliefDecisionTargetSelection {
-  id: CodeAgentBeliefDecisionTargetId;
-  label: string;
-  points: BeliefDecisionPoint[];
-  support: BeliefDecisionInventoryBucket;
-  reasons: string[];
-}
-interface SelectBeliefDecisionTargetOptions {
-  minN?: number;
-  minOutcomeCoverage?: number;
-  preferredTargets?: CodeAgentBeliefDecisionTargetId[];
-}
-interface AnalyzeBeliefDecisionCorpusOptions {
-  points: BeliefDecisionPoint[];
-  targetId?: CodeAgentBeliefDecisionTargetId;
-  minN?: number;
-  minOutcomeCoverage?: number;
-  minAccepted?: number;
-  confidenceThreshold?: number;
-  policy?: BeliefSelectivePolicy;
-  requireOpe?: boolean;
-  policyOptions?: Partial<AnalyzeBeliefPolicyOptions>;
-}
-interface BeliefDecisionCorpusEvaluation {
-  inventory: BeliefDecisionInventoryReport;
-  target?: BeliefDecisionTargetSelection;
-  policy?: BeliefSelectivePolicy;
-  evaluation?: BeliefPolicyEvaluationReport;
-  diagnostics: string[];
-}
-declare function extractCodeAgentBeliefDecisionPoints(options: ExtractCodeAgentBeliefDecisionPointsOptions): BeliefDecisionExtractionReport;
-declare function inventoryBeliefDecisionPoints(points: BeliefDecisionPoint[]): BeliefDecisionInventoryReport;
-declare function selectBeliefDecisionTarget(points: BeliefDecisionPoint[], options?: SelectBeliefDecisionTargetOptions): BeliefDecisionTargetSelection | null;
-declare function analyzeBeliefDecisionCorpus(options: AnalyzeBeliefDecisionCorpusOptions): BeliefDecisionCorpusEvaluation;
-//#endregion
-//#region src/belief-state/research-evidence.d.ts
-type BeliefResearchClaimScope = 'selective' | 'counterfactual';
-type BeliefResearchEvidenceStatus = 'supported' | 'blocked';
-type BeliefResearchGateId = 'corpus' | 'selective' | 'calibration' | 'ope';
-interface BeliefResearchEvidenceGate {
-  id: BeliefResearchGateId;
-  status: BeliefResearchEvidenceStatus;
-  blockers: string[];
-  caveats: string[];
-}
-interface BeliefDecisionResearchEvidencePacket {
-  claimScope: BeliefResearchClaimScope;
-  status: BeliefResearchEvidenceStatus;
-  analysis: BeliefDecisionCorpusEvaluation;
-  gates: BeliefResearchEvidenceGate[];
-  blockers: string[];
-  caveats: string[];
-}
-interface BuildBeliefDecisionResearchEvidencePacketOptions extends AnalyzeBeliefDecisionCorpusOptions {
-  claimScope?: BeliefResearchClaimScope;
-}
-declare function buildBeliefDecisionResearchEvidencePacket(options: BuildBeliefDecisionResearchEvidencePacketOptions): BeliefDecisionResearchEvidencePacket;
-//#endregion
-//#region src/belief-state/code-agent-evidence.d.ts
-interface CodeAgentBeliefSession extends CodeAgentSessionIntakeOptions {
-  source: CodeAgentSessionSource;
-}
-interface BuildCodeAgentBeliefEvidenceCorpusOptions extends Omit<BuildBeliefDecisionResearchEvidencePacketOptions, 'points'> {
-  sessions: CodeAgentBeliefSession[];
-}
-interface CodeAgentBeliefEvidenceCorpus {
-  runs: RunRecord[];
-  metrics: CodeAgentSessionMetrics[];
-  intakeDiagnostics: CodeAgentSessionDiagnostic[];
-  extractionDiagnostics: BeliefDecisionExtractionDiagnostic[];
-  decisions: BeliefDecisionPoint[];
-  inventory: BeliefDecisionInventoryReport;
-  evidence: BeliefDecisionResearchEvidencePacket;
-}
-declare function buildCodeAgentBeliefEvidenceCorpus(options: BuildCodeAgentBeliefEvidenceCorpusOptions): CodeAgentBeliefEvidenceCorpus;
-//#endregion
-//#region src/belief-state/extract.d.ts
-interface ExtractBeliefDecisionPointsOptions {
-  runIds?: string[];
-}
-declare function extractBeliefDecisionPoints(store: TraceStore, options?: ExtractBeliefDecisionPointsOptions): Promise<BeliefDecisionExtractionReport>;
-//#endregion
-//#region src/belief-state/shadow-probe.d.ts
-interface BeliefShadowProbeInput {
-  probeId: string;
-  decisionId: string;
-  runId: string;
-  scenarioId?: string;
-  stepIndex: number;
-  decisionKind: BeliefDecisionKind;
-  candidateActions: string[];
-  observedAction?: string;
-  evidence: BeliefShadowProbeEvidenceRef[];
-  context?: string;
-  metadata?: Record<string, unknown>;
-}
-interface BeliefShadowProbeEvidenceRef {
-  id: string;
-  source: string;
-  detail?: string;
-  quality?: BeliefEvidenceQuality;
-}
-interface BeliefShadowProbeResponse {
-  predictedAction: string;
-  confidence: number;
-  beliefSummary?: string;
-  uncertainty?: string[];
-  evidenceRefs?: string[];
-  wouldChangeMindIf?: string[];
-  targetProb?: number;
-  qHatChosen?: number | null;
-  vHatTarget?: number | null;
-  metadata?: Record<string, unknown>;
-}
-interface BeliefShadowProbeRecord extends BeliefShadowProbeResponse {
-  probeId: string;
-  decisionId: string;
-  runId: string;
-  scenarioId?: string;
-  stepIndex: number;
-  decisionKind: BeliefDecisionKind;
-  candidateActions: string[];
-  observedAction: string;
-  agreesWithObservedAction: boolean;
-  outcome?: BeliefDecisionOutcome;
-}
-interface BeliefShadowProbeDiagnostic {
-  decisionId: string;
-  severity: 'warning' | 'error';
-  reason: string;
-}
-interface BeliefShadowProbeSummary {
-  attempted: number;
-  completed: number;
-  dropped: number;
-  withOutcome: number;
-  withTargetProb: number;
-  meanConfidence: number | null;
-  observedAgreementRate: number | null;
-}
-interface BeliefShadowProbeRun {
-  probeId: string;
-  records: BeliefShadowProbeRecord[];
-  diagnostics: BeliefShadowProbeDiagnostic[];
-  summary: BeliefShadowProbeSummary;
-}
-interface RunBeliefShadowProbeOptions {
-  probeId: string;
-  points: BeliefDecisionPoint[];
-  probe: (input: BeliefShadowProbeInput) => BeliefShadowProbeResponse | Promise<BeliefShadowProbeResponse>;
-  contextOf?: (point: BeliefDecisionPoint) => string | undefined | Promise<string | undefined>;
-  metadataOf?: (point: BeliefDecisionPoint) => Record<string, unknown> | undefined | Promise<Record<string, unknown> | undefined>;
-  includeObservedAction?: boolean;
-  includeEvidenceDetail?: boolean;
-  includeOutcomeInRecord?: boolean;
-  requireCandidateActions?: boolean;
-  allowOutOfSetActions?: boolean;
-  concurrency?: number;
-  maxContextChars?: number;
-}
-declare function runBeliefShadowProbe(options: RunBeliefShadowProbeOptions): Promise<BeliefShadowProbeRun>;
-declare function formatBeliefShadowProbePrompt(input: BeliefShadowProbeInput): string;
-//#endregion
-//#region src/belief-state/runtime-hooks.d.ts
-interface RuntimeBeliefDecisionEvidenceRef {
-  source: string;
-  id: string;
-  detail?: string;
-  quality?: BeliefEvidenceQuality;
-  metadata?: Record<string, unknown>;
-}
-interface RuntimeBeliefDecisionPoint {
-  id: string;
-  runId: string;
-  scenarioId?: string;
-  stepIndex: number;
-  kind: string;
-  candidateActions?: string[];
-  context?: string;
-  evidence?: RuntimeBeliefDecisionEvidenceRef[];
-  metadata?: Record<string, unknown>;
-}
-interface RuntimeBeliefHookEvent {
-  id: string;
-  runId: string;
-  scenarioId?: string;
-  target: string;
-  phase: string;
-  timestamp: number;
-  stepIndex?: number;
-  parentId?: string;
-  payload?: unknown;
-  metadata?: Record<string, unknown>;
-}
-interface RuntimeBeliefHookContext {
-  signal?: AbortSignal;
-}
-interface RuntimeBeliefHooks {
-  onEvent?: (event: RuntimeBeliefHookEvent, context: RuntimeBeliefHookContext) => void | Promise<void>;
-  onDecisionPoint?: (point: RuntimeBeliefDecisionPoint, context: RuntimeBeliefHookContext) => void | Promise<void>;
-}
-interface RuntimeBeliefConversionDiagnostic {
-  decisionId: string;
-  severity: 'warning' | 'error';
-  reason: string;
-}
-interface RuntimeBeliefShadowProbeInputOptions {
-  probeId: string;
-  decisionKind?: BeliefDecisionKind;
-  includeEvidenceDetail?: boolean;
-  includeLifecycleEvidence?: boolean;
-  lifecycleEvents?: RuntimeBeliefHookEvent[];
-  maxContextChars?: number;
-}
-interface RuntimeBeliefDecisionPointOptions {
-  chosenAction?: string;
-  decisionKind?: BeliefDecisionKind;
-  confidence?: number;
-  behaviorProb?: number;
-  targetProb?: number;
-  qHatChosen?: number | null;
-  vHatTarget?: number | null;
-  costUsd?: number;
-  outcome?: BeliefDecisionOutcome;
-  metadata?: Record<string, unknown>;
-  includeLifecycleEvidence?: boolean;
-  lifecycleEvents?: RuntimeBeliefHookEvent[];
-}
-interface RuntimeBeliefShadowProbeInputReport {
-  input?: BeliefShadowProbeInput;
-  diagnostics: RuntimeBeliefConversionDiagnostic[];
-}
-interface RuntimeBeliefDecisionPointReport {
-  point?: BeliefDecisionPoint;
-  diagnostics: RuntimeBeliefConversionDiagnostic[];
-}
-interface BeliefRuntimeHookCollector {
-  hooks: RuntimeBeliefHooks;
-  decisions: RuntimeBeliefDecisionPoint[];
-  events: RuntimeBeliefHookEvent[];
-  toShadowProbeInputs(options?: Partial<RuntimeBeliefShadowProbeInputOptions>): {
-    inputs: BeliefShadowProbeInput[];
-    diagnostics: RuntimeBeliefConversionDiagnostic[];
-  };
-  clear(): void;
-}
-declare function runtimeDecisionPointToBeliefShadowProbeInput(point: RuntimeBeliefDecisionPoint, options: RuntimeBeliefShadowProbeInputOptions): RuntimeBeliefShadowProbeInputReport;
-declare function runtimeDecisionPointToBeliefDecisionPoint(point: RuntimeBeliefDecisionPoint, options: RuntimeBeliefDecisionPointOptions): RuntimeBeliefDecisionPointReport;
-declare function createBeliefRuntimeHookCollector(defaults: RuntimeBeliefShadowProbeInputOptions): BeliefRuntimeHookCollector;
-//#endregion
-//#region src/belief-state/phase0-measurement.d.ts
-interface RuntimeBeliefPhase0RunRecord {
-  runId: string;
-  scenarioId?: string;
-  splitTag: RunSplitTag;
-}
-interface RuntimeBeliefDecisionLabel {
-  decisionId: string;
-  chosenAction: string;
-  outcome: BeliefDecisionOutcome;
-  confidence?: number;
-  behaviorProb?: number;
-  targetProb?: number;
-  qHatChosen?: number | null;
-  vHatTarget?: number | null;
-  costUsd?: number;
-  splitTag?: RunSplitTag;
-  metadata?: Record<string, unknown>;
-}
-interface BuildRuntimeBeliefPhase0MeasurementOptions extends Omit<BuildBeliefDecisionResearchEvidencePacketOptions, 'points'> {
-  runs: RuntimeBeliefPhase0RunRecord[];
-  decisions: RuntimeBeliefDecisionPoint[];
-  events?: RuntimeBeliefHookEvent[];
-  labels: RuntimeBeliefDecisionLabel[];
-  baselinePolicyId?: string;
-}
-interface RuntimeBeliefPhase0MeasurementSummary {
-  runCount: number;
-  producerDecisionCount: number;
-  lifecycleEventCount: number;
-  labelCount: number;
-  completedPointCount: number;
-  runJoinRate: number;
-  labelJoinRate: number;
-  missingRunRecordCount: number;
-  missingLabelCount: number;
-  withEvidence: number;
-  withOutcome: number;
-  withSplit: number;
-  withBehaviorProb: number;
-  withTargetProb: number;
-  baselinePolicyId: string;
-  packetStatus: BeliefDecisionResearchEvidencePacket['status'];
-  claimScope: BeliefDecisionResearchEvidencePacket['claimScope'];
-}
-interface RuntimeBeliefPhase0Measurement {
-  points: BeliefDecisionPoint[];
-  packet: BeliefDecisionResearchEvidencePacket;
-  summary: RuntimeBeliefPhase0MeasurementSummary;
-  diagnostics: string[];
-}
-declare function buildRuntimeBeliefPhase0Measurement(options: BuildRuntimeBeliefPhase0MeasurementOptions): RuntimeBeliefPhase0Measurement;
-//#endregion
-//#region src/belief-state/runtime-benchmark-corpus.d.ts
-type RuntimeBenchmarkTrajectoryRecord = RuntimeTrajectoryRecord & {
-  benchmark?: unknown;
-  condition?: unknown;
-  instanceId?: unknown;
-  runtimeDecisionPoints?: unknown;
-};
-interface BuildRuntimeBenchmarkBeliefPhase0MeasurementOptions extends Omit<BuildRuntimeBeliefPhase0MeasurementOptions, 'runs' | 'events' | 'decisions' | 'labels'> {
-  records: RuntimeBenchmarkTrajectoryRecord[];
-  decisions?: RuntimeBeliefDecisionPoint[];
-  defaultSplitTag?: ProjectRuntimeTrajectoryEvidenceOptions['defaultSplitTag'];
-  labels?: RuntimeBeliefDecisionLabel[];
-}
-interface RuntimeBenchmarkBeliefPhase0Summary {
-  decisionCount: number;
-  labelCount: number;
-}
-interface RuntimeBenchmarkBeliefPhase0Measurement {
-  runs: RuntimeBeliefPhase0RunRecord[];
-  events: RuntimeBeliefHookEvent[];
-  decisions: RuntimeBeliefDecisionPoint[];
-  labels: RuntimeBeliefDecisionLabel[];
-  trajectory: RuntimeTrajectoryEvidenceProjection;
-  measurement: RuntimeBeliefPhase0Measurement;
-  summary: RuntimeBenchmarkBeliefPhase0Summary;
-  diagnostics: string[];
-}
-declare function buildRuntimeBenchmarkBeliefPhase0Measurement(options: BuildRuntimeBenchmarkBeliefPhase0MeasurementOptions): RuntimeBenchmarkBeliefPhase0Measurement;
-//#endregion
-export { AnalyzeBeliefDecisionCorpusOptions, AnalyzeBeliefPolicyOpeOptions, AnalyzeBeliefPolicyOptions, BELIEF_DECISION_KINDS, BELIEF_EVALUATION_CRITERIA, BELIEF_EVIDENCE_QUALITIES, BELIEF_EVIDENCE_SOURCES, BeliefCalibrationOptions, BeliefCalibrationRegion, BeliefCalibrationStatus, BeliefDecisionCorpusEvaluation, BeliefDecisionExtractionDiagnostic, BeliefDecisionExtractionReport, BeliefDecisionInventoryBucket, BeliefDecisionInventoryReport, BeliefDecisionKind, BeliefDecisionOutcome, BeliefDecisionPoint, BeliefDecisionReason, BeliefDecisionReasonCode, BeliefDecisionResearchEvidencePacket, BeliefDecisionTargetSelection, BeliefEvaluationCriterionId, BeliefEvaluationStatus, BeliefEvidenceQuality, BeliefEvidenceRef, BeliefEvidenceSource, BeliefOffPolicyTrajectoryReport, BeliefOpeOptions, BeliefOpeReport, BeliefOpeStatus, BeliefOpeSupportDiagnostics, BeliefOpeTargetPolicy, BeliefPolicyAction, BeliefPolicyDecision, BeliefPolicyEvaluationReport, BeliefResearchClaimScope, BeliefResearchEvidenceGate, BeliefResearchEvidenceStatus, BeliefResearchGateId, BeliefRuntimeHookCollector, BeliefSelectivePolicy, BeliefSelectivePolicyMetrics, BeliefShadowProbeDiagnostic, BeliefShadowProbeEvidenceRef, BeliefShadowProbeInput, BeliefShadowProbeRecord, BeliefShadowProbeResponse, BeliefShadowProbeRun, BeliefShadowProbeSummary, BeliefUtilityOptions, BuildBeliefDecisionResearchEvidencePacketOptions, BuildCodeAgentBeliefEvidenceCorpusOptions, BuildRuntimeBeliefPhase0MeasurementOptions, BuildRuntimeBenchmarkBeliefPhase0MeasurementOptions, CodeAgentBeliefDecisionTargetId, CodeAgentBeliefEvidenceCorpus, CodeAgentBeliefSession, EvaluateBeliefSelectivePolicyOptions, ExtractBeliefDecisionPointsOptions, ExtractCodeAgentBeliefDecisionPointsOptions, RunBeliefShadowProbeOptions, RuntimeBeliefConversionDiagnostic, RuntimeBeliefDecisionEvidenceRef, RuntimeBeliefDecisionLabel, RuntimeBeliefDecisionPoint, RuntimeBeliefDecisionPointOptions, RuntimeBeliefDecisionPointReport, RuntimeBeliefHookContext, RuntimeBeliefHookEvent, RuntimeBeliefHooks, RuntimeBeliefPhase0Measurement, RuntimeBeliefPhase0MeasurementSummary, RuntimeBeliefPhase0RunRecord, RuntimeBeliefShadowProbeInputOptions, RuntimeBeliefShadowProbeInputReport, RuntimeBenchmarkBeliefPhase0Measurement, RuntimeBenchmarkBeliefPhase0Summary, SelectBeliefDecisionTargetOptions, analyzeBeliefDecisionCorpus, analyzeBeliefPolicy, beliefDecisionsToOffPolicyTrajectories, buildBeliefDecisionResearchEvidencePacket, buildCodeAgentBeliefEvidenceCorpus, buildRuntimeBeliefPhase0Measurement, buildRuntimeBenchmarkBeliefPhase0Measurement, calibrateBeliefDecisions, createBeliefRuntimeHookCollector, embeddedBeliefOpeTargetPolicy, evaluateBeliefOffPolicy, evaluateBeliefSelectivePolicy, extractBeliefDecisionPoints, extractCodeAgentBeliefDecisionPoints, formatBeliefShadowProbePrompt, inventoryBeliefDecisionPoints, isBeliefDecisionKind, isBeliefEvidenceSource, runBeliefShadowProbe, runtimeDecisionPointToBeliefDecisionPoint, runtimeDecisionPointToBeliefShadowProbeInput, selectBeliefDecisionTarget, thresholdSelectivePolicy };
-//# sourceMappingURL=index.d.ts.map

package/dist/belief-state/index.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.d.ts","names":[],"sources":["../../src/belief-state/types.ts","../../src/belief-state/calibration.ts","../../src/belief-state/ope.ts","../../src/belief-state/selective.ts","../../src/belief-state/report.ts","../../src/belief-state/code-agent-corpus.ts","../../src/belief-state/research-evidence.ts","../../src/belief-state/code-agent-evidence.ts","../../src/belief-state/extract.ts","../../src/belief-state/shadow-probe.ts","../../src/belief-state/runtime-hooks.ts","../../src/belief-state/phase0-measurement.ts","../../src/belief-state/runtime-benchmark-corpus.ts"],"mappings":";;;;;;;cAGa;KAcD,6BAA6B;cAE5B;KAUD,+BAA+B;cAE9B;KASD,gCAAgC;cAE/B;WAET;WACA;WACA;;WAGA;WACA;WACA;;WAOA;WACA;WACA;;WAQA;WACA;WACA;;WAGA;WACA;WACA;;WAGA;WACA;WACA;;WAGA;WACA;WACA;;WAGA;WACA;WACA;;WAUA;WACA;WACA;;WAQA;WACA;WACA;;WAGA;WACA;WACA;;WAQA;WACA;WACA;;KAIQ,sCAAsC;KACtC,mCACF;UAEO;EACf,MAAM;EACN,YAAY;EACZ;EACA;EACA,WAAW;;iBAGG,qBAAqB,iBAAiB,SAAS;iBAI/C,uBAAuB,iBAAiB,SAAS;UAMhD;EACf,QAAQ;EACR;EACA;EACA;EACA;EACA;EACA,UAAU;EACV;EACA,WAAW;;UAGI;EACf;EACA;EACA;EACA;EACA;EACA,WAAW;;UAGI;EACf;EACA;EACA;EACA;EACA,MAAM;EACN;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,UAAU;EACV,UAAU;EACV,UAAU;EACV,WAAW;;UAGI;EACf;EACA;EACA;EACA;;UAGe;EACf,WAAW;EACX,aAAa;;KAGH;UAEK;EACf,QAAQ;EACR;EACA;EACA;EACA;EACA;EACA,UAAU;;UAGK;EACf;EACA,OAAO,OAAO,sBAAsB;;UAGrB;EACf;EACA,aAAa,OAAO;EACpB,cAAc,OAAO;EACrB,cAAc,OAAO;;UAGN;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;IAAe;IAAc;IAAe;;EAC5C;EACA;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf;EACA,KAAK;EACL,OAAO;EACP,IAAI;EACJ,SAAS;;KAGC;KACA;KACA;UAEK;EACf;EACA;EACA,QAAQ;EACR,iBAAiB;EACjB,mBAAmB;EACnB,WAAW;EACX;EACA,WAAW;EACX,cAAc;EACd,MAAM;EACN;;;;KC9RU;UAEK;EACf;EACA;EACA,SAAS;EACT,SAAS;;iBAGK,yBACd,QAAQ,uBACR,UAAS,2BAAA;;;UCHM,yBAAyB;EACxC;EACA;EACA;;UAGe;EACf;EACA,cAAc;EACd;EACA;;iBAGc,8BAA8B,cAA8B;iBAe5D,uCACd,QAAQ,uBACR,cAAc,uBACd,UAAS,KAAK,sCACb;iBA+Da,wBACd,QAAQ,uBACR,cAAc,uBACd,UAAS,mBACR;;;UCrGc;EACf,UAAU;EACV;EACA;EACA;EACA;;iBAcc,yBAAyB;EACvC;EACA;EACA,uBAAuB,QAAQ;IAC7B;iBAyBY,8BACd,QAAQ,uBACR,QAAQ,uBACR,UAAS,uCACR;;;UC/Cc,sCAAsC;EACrD,eAAe;;UAGA;EACf,QAAQ;EACR,QAAQ;EACR,YAAY;EACZ,cAAc;EACd,MAAM;EACN;;iBAGc,oBACd,SAAS,6BACR;;;KCRS;UAKK;EACf,QAAQ;EACR;;EAEA,cAAc;EACd,KAAK,KAAK;EACV;;UAGe;EACf;EACA,OAAO;EACP,WAAW;EACX;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf;EACA,QAAQ;EACR,UAAU;EACV;;UAGe;EACf,IAAI;EACJ;EACA,QAAQ;EACR,SAAS;EACT;;UAGe;EACf;EACA;EACA,mBAAmB;;UAGJ;EACf,QAAQ;EACR,WAAW;EACX;EACA;EACA;EACA;EACA,SAAS;EACT;EACA,gBAAgB,QAAQ;;UAGT;EACf,WAAW;EACX,SAAS;EACT,SAAS;EACT,aAAa;EACb;;iBA0Bc,qCACd,SAAS,8CACR;iBA4Ca,8BACd,QAAQ,wBACP;iBA2Ba,2BACd,QAAQ,uBACR,UAAS,oCACR;iBAmCa,4BACd,SAAS,qCACR;;;KChOS;KACA;KACA;UAEK;EACf,IAAI;EACJ,QAAQ;EACR;EACA;;UAGe;EACf,YAAY;EACZ,QAAQ;EACR,UAAU;EACV,OAAO;EACP;EACA;;UAGe,yDACP;EACR,aAAa;;iBAGC,0CACd,SAAS,mDACR;;;UCRc,+BAA+B;EAC9C,QAAQ;;UAGO,kDACP,KAAK;EACb,UAAU;;UAGK;EACf,MAAM;EACN,SAAS;EACT,mBAAmB;EACnB,uBAAuB;EACvB,WAAW;EACX,WAAW;EACX,UAAU;;iBAGI,mCACd,SAAS,4CACR;;;UCnCc;EACf;;iBAKoB,4BACpB,OAAO,YACP,UAAS,qCACR,QAAQ;;;UCbM;EACf;EACA;EACA;EACA;EACA;EACA,cAAc;EACd;EACA;EACA,UAAU;EACV;EACA,WAAW;;UAGI;EACf;EACA;EACA;EACA,UAAU;;UAGK;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;;UAGI,gCAAgC;EAC/C;EACA;EACA;EACA;EACA;EACA,cAAc;EACd;EACA;EACA;EACA,UAAU;;UAGK;EACf;EACA;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf;EACA,SAAS;EACT,aAAa;EACb,SAAS;;UAGM;EACf;EACA,QAAQ;EACR,QACE,OAAO,2BACJ,4BAA4B,QAAQ;EACzC,aAAa,OAAO,6CAA6C;EACjE,cACE,OAAO,wBACJ,sCAAsC,QAAQ;EACnD;EACA;EACA;EACA;EACA;EACA;EACA;;iBAMoB,qBACpB,SAAS,8BACR,QAAQ;iBA6BK,8BAA8B,OAAO;;;UCxHpC;EACf;EACA;EACA;EACA,UAAU;EACV,WAAW;;UAGI;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX,WAAW;;UAGI;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;;UAGI;EACf,SAAS;;UAGM;EACf,WACE,OAAO,wBACP,SAAS,oCACC;EACZ,mBACE,OAAO,4BACP,SAAS,oCACC;;UAGG;EACf;EACA;EACA;;UAGe;EACf;EACA,eAAe;EACf;EACA;EACA,kBAAkB;EAClB;;UAGe;EACf;EACA,eAAe;EACf;EACA;EACA;EACA;EACA;EACA;EACA,UAAU;EACV,WAAW;EACX;EACA,kBAAkB;;UAGH;EACf,QAAQ;EACR,aAAa;;UAGE;EACf,QAAQ;EACR,aAAa;;UAGE;EACf,OAAO;EACP,WAAW;EACX,QAAQ;EACR,oBAAoB,UAAU,QAAQ;IACpC,QAAQ;IACR,aAAa;;EAEf;;iBAMc,6CACd,OAAO,4BACP,SAAS,uCACR;iBA6Ba,0CACd,OAAO,4BACP,SAAS,oCACR;iBAsDa,iCACd,UAAU,uCACT;;;UC9Lc;EACf;EACA;EACA,UAAU;;UAGK;EACf;EACA;EACA,SAAS;EACT;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX,WAAW;;UAGI,mDACP,KAAK;EACb,MAAM;EACN,WAAW;EACX,SAAS;EACT,QAAQ;EACR;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,cAAc;EACd,YAAY;;UAGG;EACf,QAAQ;EACR,QAAQ;EACR,SAAS;EACT;;iBAKc,oCACd,SAAS,6CACR;;;KC1DE,mCAAmC;EACtC;EACA;EACA;EACA;;UAoBe,4DACP,KACN;EAGF,SAAS;EACT,YAAY;EACZ,kBAAkB;EAClB,SAAS;;UAGM;EACf;EACA;;UAGe;EACf,MAAM;EACN,QAAQ;EACR,WAAW;EACX,QAAQ;EACR,YAAY;EACZ,aAAa;EACb,SAAS;EACT;;iBAGc,6CACd,SAAS,sDACR"}