@tangle-network/agent-eval 0.221.5 → 0.221.7

package/dist/index-CfJULcW5.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index-CfJULcW5.d.ts","names":[],"sources":["../src/statistics/agreement-irr.ts","../src/meta-eval/correlation-study.ts","../src/meta-eval/calibration.ts","../src/meta-eval/cluster-bootstrap.ts","../src/meta-eval/evaluator-admission.ts","../src/series-convergence.ts","../src/meta-eval/sentinel.ts","../src/meta-eval/judge-gate.ts","../src/meta-eval/judge-panel-calibration.ts","../src/meta-eval/plants.ts"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;iBAyBgB,sBAAsB,aAAa;UA0ElC;;EAEf;;EAEA;;EAEA;;EAEA;;UAGe,oCAAoC;EACnD;;EAEA;;EAEA;;UAGe;;EAEf,cAAc;;EAEd;;EAEA;;EAEA;;EAEA;;UAGe,+BAA+B;;;;;;EAM9C;;;;;;EAMA;;;;;;;;;;;;;;;;;;iBAmBc,0BACd,SAAS,qBACT,OAAM,yBACL;;;;;;;;;iBA0Ha,yCACd,aAAa;EAAQ;EAAgB,QAAQ;IAC7C,OAAM,yBACL;;;UC9Qc;EACf;;;EAGA,WAAW,KAAK,KAAK,OAAO,eAAe;;UAG5B;EACf;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;;EAEA;IAAe;IAAe;;;EAE9B;IAAgB;IAAe;;;EAE/B;;UAGe;EACf,OAAO;;EAEP,eAAe,MACb;IACE;IACA;;EAGJ;EACA;;UAGe;;EAEf;;EAEA,gBAAgB;;EAEhB;;EAEA;;;EAGA;;iBAGoB,iBACpB,YAAY,YACZ,cAAc,cACd,aAAa,kBACb,8BACA,UAAS,0BACR,QAAQ;;;UChEM;EACf;EACA;EACA;EACA;EACA;;EAEA;;UAGe;EACf;EACA;EACA;EACA,MAAM;;EAEN;;EAEA;;UAGe;;EAEf;;EAEA;;EAEA;IAAU;IAAY;;;UAGP;EACf;EACA;;iBAGoB,iBACpB,YAAY,YACZ,cAAc,cACd,YAAY,gBACZ,uBACA,UAAS,qBACR,QAAQ;;iBA0CK,qBACd,qBAAqB,mBACrB,oBACA,uBACA,UAAS,qBACR;;;;;;iBA+Da,WAAW,gBAAgB;;UAkB1B;;EAEf,eAAe,SAAS;EACxB;;;;;;;;;iBAUc,qBAAqB,oBAAoB;;UAaxC;EACf;;EAEA;;;;;;;;iBASc,uBAAuB,oBAAoB;;;;UC1N1C;EACf;EACA;;UAGe;;EAEf;;EAEA;;EAEA;;UAGe;EACf;EACA;;EAEA;EACA;EACA;EACA;EACA;;;;;;;;;;;;iBAac,qBACd,uBAAuB,wBACvB,UAAS,0BACR;;;cCjCU,gCAA8B,EAAA;;;;GAMhC,EAAA,KAAA;;KAGC,2BAA2B,EAAE,aAAa;cAEzC,iCAA+B,EAAA;;;;;;;;;;;;;;;;;GASjC,EAAA,KAAA;;KAGC,4BAA4B,EAAE,aAAa;cAEjD,aAAW,EAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAeN,EAAA,KAAA;KAEC,sBAAsB,EAAE,aAAa;UAEhC;;EAEf;EACA;;EAEA;;EAEA;EACA;;EAEA;;EAEA,UAAU;EACV;EACA;;UAGe;EACf,iBAAiB;EACjB,cAAc;EACd,aAAa;EACb,cAAc;EACd;EACA;EACA,WAAW;EACX,QAAQ;;EAER;EACA;EACA;EACA;EACA,cAAc;EACd;IACE;IACA;IACA;IACA;IACA;IACA;;EAEF,YAAY;IAAQ;IAAY;IAA2B;;EAC3D,iBAAiB;EACjB,gBAAgB;;cAkGZ,gBAAc,EAAA;;;;;;;;;;;;;;;;;;GAOT,EAAA,KAAA;KAEC,0BAA0B,EAAE,aAAa;;;;;;iBA0BrC,mBAAmB,OAAO;;;;;;;;;;;;;;;;;KAiC9B,qBAAqB,kBAAkB;;;;;;iBAOnC,eAAe,OAAO,sBAAsB;cAoFtD,wBAAsB,EAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAG1B,EAAA,KAAA;;KAGU,8BAA8B,EAAE,aAAa;UAExC,qCAAqC;EACpD;IACE,uBAAuB;IACvB,YAAY;;IAEZ,cAAc;MAAQ;MAAY;;;;;;;;;iBAStB,uBACd,OAAO,8BACN;;;;;;;;;;;;;UC7Wc;;EAEf;;EAEA;;EAEA;;UAGe;EACf;EACA;EACA;;EAEA;;EAEA;;iBAGc,cACd,kBACA,UAAS,2BACR;;;cCoBG;KAEM,6BAA6B;UAExB;;EAEf;;EAEA;;EAEA;;UAGe;;EAEf;EACA;;EAEA;;;;;;;EAOA,SAAS;;;UAIM;EACf;EACA;EACA;;;iBAcc,yBAAyB,UAAU,kBAAkB;;;;;;iBAiCrD,wBACd,QAAQ,oBAAoB,6BAC5B,MAAM,eACL;;;;;;;;iBAmBa,sBACd,QAAQ,wBAAwB,qBAChC,MAAM,eACL;UAYc;;EAEf;;;;;;;;;iBAUc,wBACd,QAAQ,kBACR,QAAQ,cACR,MAAM,cACN,UAAS,qBACR;UA4Cc;EACf,OAAO,UAAU,mBAAmB;;EAEpC,QAAQ,mBAAmB,QAAQ;;;;;;;;iBASrB,kBAAkB,eAAe;UA0ChC;;EAEf;;EAEA;;EAEA;;EAEA;;UAGe;;EAEf;EACA,aAAa;;EAEb,cAAc;;UAGC;EACf;EACA,QAAQ;;EAER,OAAO;;EAEP;;EAEA;;EAEA;EACA;EACA;;UAGe;EACf,UAAU;EACV;EACA;;;;;;;EAOA;;iBAKc,oBACd,SAAS,oBACT,MAAM,uBACL;UA6Gc;EACf;EACA;;;;;;;;;;iBAWc,gBAAgB,QAAQ,iBAAiB;;;KCja7C;;UAGK;;EAEf;EACA,SAAS;;EAET;;EAEA;;EAEA;;EAEA;EACA;;UAGe;EACf;;EAEA;EACA,mBAAmB;;;;;;iBAuDL,uBAAuB,KAAK,sBAAsB;UAwBjD;EACf,KAAK;;EAEL;;EAEA;;EAEA;;;;;;;;iBASc,uBACd,KAAK,qBACL,mBAAmB,uBAClB;;UAkCc;EACf;;EAEA,SAAS;;EAET,mBAAmB;;UAGJ;EACf;;EAEA;EACA;EACA;;UAGe;;EAEf;;EAEA;;EAEA;;EAEA;;UAGe;EACf;EACA,SAAS;EACT;;EAEA;;EAEA;;UAGe,uBAAuB;EACtC;;EAEA;;EAEA;;EAEA,iBAAiB,OAAO;EACxB;;EAEA;;;;;;EAMA;IAAmB;IAAgB;IAAgB,UAAU;;;;;;EAK7D;;EAEA;;EAEA;EACA,WAAW;EACX;;IAEE;;IAEA;;IAEA;;;EAGF,eAAe;;EAEf;;;;;;;iBAyBc,sBACd,KAAK,qBACL,eAAe,YACf,MAAM,eACL;UA+Gc;;;;;EAKf;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA,QAAQ;;cAGG,2BAA2B,SAAS;KAUrC;UAEK;EACf;EACA;EACA;;EAEA;EACA,QAAQ;;EAER;EACA,WAAW;EACX,QAAQ;;UAGO;EACf;;EAEA;;EAEA;;EAEA,WAAW;EACX;;EAEA,mBAAmB;EACnB,SAAS,QAAQ;;;iBAUH,kBAAkB,OAAO,iBAAiB;;cA2G7C,8BAA8B;WAChC,UAAU;EAEnB,YAAY,UAAU;;;iBASR,mBAAmB,UAAU;;;;;;;;iBAW7B,2BACd,WAAW,gBACX;EAAW,UAAU;IACpB;;;;UCrkBc;EACf;EACA;;UAGe,yBAAyB;EACxC;;EAEA;;EAEA;;EAEA;;;EAGA,YAAY,QAAQ,OAAO;;UAGZ;;EAEf;;EAEA;;EAEA;;EAEA;;EAEA;;UAGe,yBAAyB;EACxC,QAAQ,SAAS,OAAO,GAAG;;EAE3B,sBAAsB;EACtB,gBAAgB,yBAAyB;EACzC;;EAEA,KAAK;EACL,QAAQ;;EAER;;EAEA;;EAEA;;EAEA,OAAO;;;;UAKQ;EACf,aAAa;EACb,QAAQ;EACR,QAAQ;;;KAIE,iBAAiB,oBAAoB,QAAQ,OAAO;EAAgB;;UAE/D,sBAAsB;EACrC;EACA,YAAY;EACZ,UAAU,eAAe,OAAO,GAAG;;EAEnC,WAAW,eAAe,OAAO,GAAG;;EAEpC,SAAS,eAAe,eAAe,iBAAiB;;EAExD,eAAe,OAAO;;EAEtB,UAAU;IAAQ;IAAe;IAAgB;;EACjD;;;iBAiDoB,oBAAoB,kBACxC,OAAO,yBAAyB,KAC/B,QAAQ,sBAAsB;;UA0GhB;;EAEf;;EAEA;;EAEA;;EAEA;;KAGU;UAEK;EACf;EACA;EACA;EACA;EACA;EACA,SAAS;EACT;;UAGe;EACf;EACA,QAAQ;EACR;EACA;EACA;EACA,OAAO;;;iBAIO,qBACd,UAAU,eAAe;EAAiB,aAAa;KACvD,QAAQ,6BACP;;;;;;;;KC1OS;;;;;;;;;;KAkBA;UAOK;;EAEf;EACA,MAAM;;EAEN,MAAM;;EAEN,iBAAiB;;;;;;;;;;;;;;iBAeH,YAAY;EAC1B;EACA,MAAM;EACN,MAAM;EACN,iBAAiB;IACf;;;;;UAqDa;;EAEf;;EAEA;;;KAIU;;KAGA;UAEK;;EAEf,UAAU;;EAEV,OAAO;;EAEP,KAAK;;;;;;EAML;;EAEA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;iBAsJc,gBAAgB,UAAU,gBAAgB;UA8BzC;EACf,OAAO;;EAEP,UAAU;EACV,OAAO;EACP,KAAK;EACL;EACA;;;;;;;;;;;;;;;;;;;;;;iBAuBc,oBAAoB;EAClC;EACA;EACA,UAAU;;EAEV;EACA;IACE;UA0Ba;;;;;;EAMf,MAAM;;EAEN;;EAEA;;EAEA;;EAEA,QAAQ;;UAGO;;;;;;EAMf,OAAO;EACP,UAAU;;UAGK;;;;;EAKf;;;;;;EAMA;;;;;;;;;;;iBAYc,WACd,kBAAkB,cAClB,iBAAiB,SACjB,UAAS,oBACR;;;;;;UAqGc;EACf;EACA;;;;;;KAOU;UAEK;EACf;EACA;EACA;EACA;;EAEA;;;;;;;;UASe;EACf;EACA;EACA;;EAEA;;UAGe;EACf,QAAQ;;EAER;EACA;EACA;EACA;;;;;EAKA;;EAEA;;EAEA,QAAQ,QAAQ,OAAO,WAAW;;EAElC;;EAEA;EACA,UAAU;;;;;;;;;;iBAWI,UACd,kBAAkB,gBAClB,UAAU,gBACT"}

package/dist/{index-y7i8l6Y_.d.ts → index-jnuFLxOs.d.ts}

@@ -1,13 +1,13 @@
 import { o as CostLedgerHandle } from "./cost-ledger-C1W3ICmP.js";
 import { a as RunRecord, c as RunSplitTag, n as RunCostProvenance } from "./run-record-C3awiXPf.js";
-import "./types-DGImvHIr.js";
-import "./statistical-heldout-pD1Yah-0.js";
-import "./index-BvX3CFNS.js";
-import "./types-DDKfH-N6.js";
+import "./types-6Lfdlshp.js";
+import "./statistical-heldout-BjzSOmLH.js";
+import "./index-CdEVr8-O.js";
+import "./types-M6eF9m-r.js";
 import { f as TraceSpanEvent } from "./client-hwo3cJjR.js";
-import { d as InsightReport, t as AnalyzeRunsOptions } from "./analyze-runs-D7In4c3z.js";
-import "./define-agent-eval-Bsoi2GUV.js";
-import "./default-registry-BMHqNR6_.js";
+import { d as InsightReport, t as AnalyzeRunsOptions } from "./analyze-runs-DfX6dHP0.js";
+import "./define-agent-eval-JKt3gMQr.js";
+import "./default-registry-CLG5YqGS.js";
 import { AgentCandidateBenchmarkCellRef, AgentCandidateBenchmarkSuiteInputs, AgentCandidateBenchmarkTask, AgentCandidateBenchmarkTaskMaterial, AgentCandidateBundle, AgentCandidateEvaluationPolicy, AgentCandidateExperiment, AgentCandidateExperimentMaterial, AgentCandidateExperimentMeasurement, AgentImprovementCost, AgentImprovementMeasuredComparison, AgentProfileImprovementExperiment, AgentProfileImprovementExperimentMaterial, AgentProfileImprovementMeasuredComparison, AgentProfileImprovementMeasurement, AgentProfileImprovementRunCell, AgentProfileImprovementRunReceipt, AgentProfileImprovementSuiteInputs, AgentProfileImprovementTask, AgentProfileImprovementTaskMaterial, CandidateExecutionEvidence, Sha256Digest } from "@tangle-network/agent-interface";
 //#region src/contract/measured-comparison-types.d.ts
 interface SealCandidateBenchmarkSuiteOptions {
@@ -678,4 +678,4 @@ interface FromOtelSpansOptions {
 declare function fromOtelSpans(opts: FromOtelSpansOptions): RunRecord[];
 //#endregion
 export { AgentProfileImprovementExperimentExecutionInput as $, CodeAgentSessionActionSurface as A, AgentTraceIndex as B, CodeAgentSessionIntakeOptions as C, PairedMeasurementAdapter as Ct, CodeAgentSessionAction as D, ParsedCodeAgentJsonl as E, SealCandidateBenchmarkSuiteOptions as Et, observeCodeAgentSession as F, parseAgentTrace as G, AgentTraceRecord as H, AgentTraceContributor as I, EvalReportingSuiteOptions as J, partitionRunsByAuthoringModel as K, AgentTraceContributorType as L, CodeAgentSessionObservation as M, CodeAgentSessionSource as N, CodeAgentSessionActionKind as O, CodeAgentSessionTerminalStatus as P, RunRecordRejection as Q, AgentTraceConversation as R, CodeAgentSessionDiagnostic as S, PairedMeasurement as St, CodeAgentSessionMetrics as T, RunCandidateExperimentOptions as Tt, AuthoringProvenance as U, AgentTraceRange as V, PartitionByAuthoringModelResult as W, FromRunRecordDirOptions as X, EvalReportingSuiteResult as Y, FromRunRecordDirResult as Z, fromPiSession as _, verifyCandidateExperimentComparison as _t, FromFeedbackTableOptions as a, runAgentProfileImprovementExperiment as at, streamCodeAgentJsonlFile as b, CompareCandidateExperimentOptions as bt, CodeAgentStoreObservation as c, sealAgentProfileImprovementTask as ct, CodeAgentStoreScope as d, measuredComparisonFromCandidateExperiment as dt, AgentProfileImprovementExperimentRun as et, CodeAgentStoreSession as f, runCandidateExperiment as ft, fromOpenCodeSession as g, verifyCandidateExperiment as gt, fromKimiCodeSession as h, sealCandidateExperiment as ht, FeedbackTableRow as i, measuredComparisonFromAgentProfileImprovementExperiment as it, CodeAgentSessionExecutionReceipt as j, CodeAgentSessionActionStatus as k, CodeAgentStoreRejection as l, verifyAgentProfileImprovementExperimentComparison as lt, fromCodexSession as m, sealCandidateBenchmarkTask as mt, fromOtelSpans as n, RunAgentProfileImprovementExperimentOptions as nt, FromFeedbackTableResult as o, sealAgentProfileImprovementExperiment as ot, fromClaudeCodeSession as p, sealCandidateBenchmarkSuite as pt, EvalReportingSuiteInput as q, FeedbackTableMeta as r, SealAgentProfileImprovementSuiteOptions as rt, fromFeedbackTable as s, sealAgentProfileImprovementSuite as st, FromOtelSpansOptions as t, CompareAgentProfileImprovementExperimentOptions as tt, CodeAgentStoreRejections as u, evaluatePairedMeasurements as ut, parseCodeAgentJsonl as v, CandidateExperimentExecutionInput as vt, CodeAgentSessionIntakeResult as w, PairedMeasurementEvaluation as wt, CodeAgentJsonlLine as x, EvaluatePairedMeasurementsOptions as xt, parseCodeAgentJsonlFile as y, CandidateExperimentRun as yt, AgentTraceFile as z };
-//# sourceMappingURL=index-y7i8l6Y_.d.ts.map
+//# sourceMappingURL=index-jnuFLxOs.d.ts.map

package/dist/{index-y7i8l6Y_.d.ts.map → index-jnuFLxOs.d.ts.map}

@@ -1 +1 @@
-{"version":3,"file":"index-y7i8l6Y_.d.ts","names":[],"sources":["../src/contract/measured-comparison-types.ts","../src/contract/candidate-experiment.ts","../src/contract/paired-measurement-evaluation.ts","../src/contract/profile-measured-comparison.ts","../src/contract/intake/run-record-dir.ts","../src/contract/eval-reporting-suite.ts","../src/contract/intake/agent-trace.ts","../src/contract/intake/code-agent-observation.ts","../src/contract/intake/code-agent-session-types.ts","../src/contract/intake/code-agent-jsonl.ts","../src/contract/intake/code-agent-session.ts","../src/contract/intake/code-agent-store.ts","../src/contract/intake/feedback-table.ts","../src/contract/intake/otel-spans.ts"],"mappings":";;;;;;;;;;;;UAeiB;EACf,QAAQ,gCAAgC;EACxC;EACA;;UAGe;EACf,YAAY;EACZ;EACA,QAAQ;EACR,MAAM;EACN,eAAe;EACf;EACA,SAAS;;UAGM;EACf,YAAY;EACZ,QAAQ,OAAO,oCAAoC,QAAQ;;EAE3D;;EAEA,aAAa;EACb,SAAS;;UAGM;EACf,cAAc;EACd;IACE;IACA,MAAM;;;UAIO;EACf,YAAY;EACZ,cAAc;EACd;IACE;IACA,MAAM;;EAER,aAAa;EACb;EACA,YAAY;EACZ;EACA,WAAW;;;UAII,kBAAkB;EACjC;EACA,UAAU;EACV,WAAW;;;UAII,yBAAyB;EACxC,MAAM,KAAK;EACX,WAAW,KAAK;IAAkB;IAAc;;EAChD,QAAQ,KAAK;EACb,eAAe,KAAK,OAAO;EAC3B,UAAU,KAAK;EACf,UAAU,KAAK;EACf,OAAO,KAAK;;UAGG,kCAAkC;EACjD,uBAAuB,kBAAkB;EACzC,QAAQ;EACR,SAAS,yBAAyB;;EAElC;;EAEA,kBAAkB;;EAElB,kBAAkB;;;KAIR,8BAA8B,KACxC;EAGA,iBAAiB;EACjB,WAAW;EACX;;;;;iBC9Dc,2BACd,UAAU,sCACT;;iBAQa,4BACd,SAAS,qCACR;;iBAiBa,wBACd,UAAU,mCACT;iBAQa,0BAA0B,iBAAiB;;iBAarC,uBACpB,SAAS,gCACR,QAAQ;;iBA4DK,0CACd,SAAS,oCACR;;iBA2Da,oCACd,iBACC;;;;;;;;;;iBC3La,2BAA2B,MACzC,SAAS,kCAAkC,QAC1C;;;UCDc;EACf,aAAa;EACb,QAAQ,gCAAgC;EACxC;EACA;;UAGe;EACf,YAAY;EACZ;EACA,aAAa;EACb,MAAM;EACN,SAAS;EACT;EACA,SAAS;;UAGM;EACf,YAAY;EACZ,QACE,OAAO,kDACN,QAAQ;;EAEX;;EAEA,aAAa;EACb,SAAS;;UAGM;EACf,cAAc;EACd;IACE;IACA,MAAM;;;UAIO;EACf,YAAY;EACZ,cAAc;EACd;IACE;IACA,MAAM;;EAER,aAAa;EACb;EACA,YAAY;EACZ;EACA,WAAW;;;iBAIG,gCACd,UAAU,sCACT;;iBAQa,iCACd,SAAS,0CACR;;iBAqBa,sCACd,UAAU,4CACT;;;;;;iBAsBmB,qCACpB,SAAS,8CACR,QAAQ;;iBA0CK,wDACd,SAAS,kDACR;;iBA2Da,kDACd,iBACC;;;;UClOc;;EAEf;;EAEA;;EAEA;;UAGe;;;;;;EAMf;;;;;;;EAOA,WAAW;;;;;;EAMX;;UAGe;;EAEf,MAAM;;EAEN,UAAU;;EAEV;;;;;;KCxBU,0BAA0B;UAErB;;;;;EAKf,UAAU,KAAK;;EAEf,OAAO;;;;;;;;;;EAUP;;;;UAKe;;;EAGf,QAAQ;;EAER;;IAEE;;IAEA;;;IAGA;;IAEA;;;IAGA,UAAU;;;EAGZ;;;;KCnDU;UAEK;EACf,MAAM;;EAEN;;UAGe;EACf;EACA;EACA;;;EAGA,cAAc;;UAGC;EACf;EACA,cAAc;EACd,QAAQ;;UAGO;EACf;EACA,eAAe;;UAGA;EACf;EACA;EACA;EACA;IAAQ;IAAc;;EACtB;IAAS;IAAe;;EACxB,OAAO;;;;UAOQ;EACf;;EAEA;;EAEA;EACA;EACA;;EAEA;;EAEA;;KAGU,kBAAkB,YAAY;;;;;;iBAW1B,gBAAgB,SAAS,qBAAqB;UAmE7C;;;;EAIf,SAAS,YAAY;;;EAGrB,cAAc;;;;;;;;iBASA,8BACd,MAAM,aACN,OAAO,kBACN;;;;;;;;KCjKS;KAQA;KAEA;KAEA;KASA;UAEK;EACf;EACA;EACA;;UAGe;EACf;EACA;EACA,MAAM;EACN,SAAS;EACT;EACA,QAAQ;EACR;EACA;EACA,UAAU;;UAGK;EACf,QAAQ;EACR;EACA;EACA;IACE,QAAQ;IACR;;EAEF,SAAS;;UAGD;EACR,QAAQ;EACR;EACA;EACA,YAAY;;;;;;;iBAeE,wBACd,SAAS,iCACR;;;UClFc;EACf;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf,QAAQ;EACR;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX;;UAGe;EACf,MAAM;EACN,aAAa;EACb,SAAS;EACT,cAAc;;UAGC;EACf;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX;EACA;EACA;EACA;EACA;EACA;;;EAGA,iBAAiB;;;EAGjB,YAAY;;;;KAKF;EACN;EAAe;EAAoB;;EACnC;EAAmB;;;;iBCvFT,oBAAoB,gBAAgB;;;;;;;;;;iBAuB7B,yBAAyB,eAAe,eAAe;;;;;iBA6BxD,wBAAwB,eAAe,QAAQ;;;iBCDrD,iBACd,SAAS,gCACR;iBAIa,sBACd,SAAS,gCACR;iBAIa,oBACd,SAAS,gCACR;iBAIa,oBACd,SAAS,gCACR;;iBAKa,cACd,SAAS,gCACR;;;;UCrDc;;;;;WAKN;;WAEA,QAAQ;;;;;WAKR;;;;;WAKA;aACE;;aAEA;;;;KAKD;;UAUK;WACN;WACA;WACA;WACA;WACA,aAAa;;;UAIP;WACN;WACA,UAAU,SAAS,QAAQ,OAAO;;WAElC,QAAQ;aACN;aACA,QAAQ;;;;;;;;;KAUT;WAEG;WACA,OAAO;WACP;WACA,mBAAmB;;WAEnB;;WAEA;;WAEA;WACA,WAAW,SAAS,QAAQ,OAAO;WACnC,YAAY;;WAGZ;WACA,OAAO;WACP;;WAEA;WACA,YAAY;;;;UCvGV;;;EAGf;;EAEA;;;EAGA;;;EAGA,WAAW;;UAGI;EACf;;;EAGA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;;EAGA,WAAW;;;EAGX,SAAS;;UAGM;;EAEf,SAAS;;;EAGT,OAAO;;;;EAIP;IAAU;IAAa;;;;;EAIvB;;UAGe;EACf,MAAM;;;EAGN,aAAa;IAAQ;IAAe;IAAe;;;iBAGrC,kBAAkB,MAAM,2BAA2B;;;UC3BlD;EACf,OAAO;;EAEP,eAAe;;EAEf;;;;;;EAMA,eAAe,eAAe,gBAAgB;;iBAGhC,cAAc,MAAM,uBAAuB"}
+{"version":3,"file":"index-jnuFLxOs.d.ts","names":[],"sources":["../src/contract/measured-comparison-types.ts","../src/contract/candidate-experiment.ts","../src/contract/paired-measurement-evaluation.ts","../src/contract/profile-measured-comparison.ts","../src/contract/intake/run-record-dir.ts","../src/contract/eval-reporting-suite.ts","../src/contract/intake/agent-trace.ts","../src/contract/intake/code-agent-observation.ts","../src/contract/intake/code-agent-session-types.ts","../src/contract/intake/code-agent-jsonl.ts","../src/contract/intake/code-agent-session.ts","../src/contract/intake/code-agent-store.ts","../src/contract/intake/feedback-table.ts","../src/contract/intake/otel-spans.ts"],"mappings":";;;;;;;;;;;;UAeiB;EACf,QAAQ,gCAAgC;EACxC;EACA;;UAGe;EACf,YAAY;EACZ;EACA,QAAQ;EACR,MAAM;EACN,eAAe;EACf;EACA,SAAS;;UAGM;EACf,YAAY;EACZ,QAAQ,OAAO,oCAAoC,QAAQ;;EAE3D;;EAEA,aAAa;EACb,SAAS;;UAGM;EACf,cAAc;EACd;IACE;IACA,MAAM;;;UAIO;EACf,YAAY;EACZ,cAAc;EACd;IACE;IACA,MAAM;;EAER,aAAa;EACb;EACA,YAAY;EACZ;EACA,WAAW;;;UAII,kBAAkB;EACjC;EACA,UAAU;EACV,WAAW;;;UAII,yBAAyB;EACxC,MAAM,KAAK;EACX,WAAW,KAAK;IAAkB;IAAc;;EAChD,QAAQ,KAAK;EACb,eAAe,KAAK,OAAO;EAC3B,UAAU,KAAK;EACf,UAAU,KAAK;EACf,OAAO,KAAK;;UAGG,kCAAkC;EACjD,uBAAuB,kBAAkB;EACzC,QAAQ;EACR,SAAS,yBAAyB;;EAElC;;EAEA,kBAAkB;;EAElB,kBAAkB;;;KAIR,8BAA8B,KACxC;EAGA,iBAAiB;EACjB,WAAW;EACX;;;;;iBC9Dc,2BACd,UAAU,sCACT;;iBAQa,4BACd,SAAS,qCACR;;iBAiBa,wBACd,UAAU,mCACT;iBAQa,0BAA0B,iBAAiB;;iBAarC,uBACpB,SAAS,gCACR,QAAQ;;iBA4DK,0CACd,SAAS,oCACR;;iBA2Da,oCACd,iBACC;;;;;;;;;;iBC3La,2BAA2B,MACzC,SAAS,kCAAkC,QAC1C;;;UCDc;EACf,aAAa;EACb,QAAQ,gCAAgC;EACxC;EACA;;UAGe;EACf,YAAY;EACZ;EACA,aAAa;EACb,MAAM;EACN,SAAS;EACT;EACA,SAAS;;UAGM;EACf,YAAY;EACZ,QACE,OAAO,kDACN,QAAQ;;EAEX;;EAEA,aAAa;EACb,SAAS;;UAGM;EACf,cAAc;EACd;IACE;IACA,MAAM;;;UAIO;EACf,YAAY;EACZ,cAAc;EACd;IACE;IACA,MAAM;;EAER,aAAa;EACb;EACA,YAAY;EACZ;EACA,WAAW;;;iBAIG,gCACd,UAAU,sCACT;;iBAQa,iCACd,SAAS,0CACR;;iBAqBa,sCACd,UAAU,4CACT;;;;;;iBAsBmB,qCACpB,SAAS,8CACR,QAAQ;;iBA0CK,wDACd,SAAS,kDACR;;iBA2Da,kDACd,iBACC;;;;UClOc;;EAEf;;EAEA;;EAEA;;UAGe;;;;;;EAMf;;;;;;;EAOA,WAAW;;;;;;EAMX;;UAGe;;EAEf,MAAM;;EAEN,UAAU;;EAEV;;;;;;KCxBU,0BAA0B;UAErB;;;;;EAKf,UAAU,KAAK;;EAEf,OAAO;;;;;;;;;;EAUP;;;;UAKe;;;EAGf,QAAQ;;EAER;;IAEE;;IAEA;;;IAGA;;IAEA;;;IAGA,UAAU;;;EAGZ;;;;KCnDU;UAEK;EACf,MAAM;;EAEN;;UAGe;EACf;EACA;EACA;;;EAGA,cAAc;;UAGC;EACf;EACA,cAAc;EACd,QAAQ;;UAGO;EACf;EACA,eAAe;;UAGA;EACf;EACA;EACA;EACA;IAAQ;IAAc;;EACtB;IAAS;IAAe;;EACxB,OAAO;;;;UAOQ;EACf;;EAEA;;EAEA;EACA;EACA;;EAEA;;EAEA;;KAGU,kBAAkB,YAAY;;;;;;iBAW1B,gBAAgB,SAAS,qBAAqB;UAmE7C;;;;EAIf,SAAS,YAAY;;;EAGrB,cAAc;;;;;;;;iBASA,8BACd,MAAM,aACN,OAAO,kBACN;;;;;;;;KCjKS;KAQA;KAEA;KAEA;KASA;UAEK;EACf;EACA;EACA;;UAGe;EACf;EACA;EACA,MAAM;EACN,SAAS;EACT;EACA,QAAQ;EACR;EACA;EACA,UAAU;;UAGK;EACf,QAAQ;EACR;EACA;EACA;IACE,QAAQ;IACR;;EAEF,SAAS;;UAGD;EACR,QAAQ;EACR;EACA;EACA,YAAY;;;;;;;iBAeE,wBACd,SAAS,iCACR;;;UClFc;EACf;EACA;;UAGe;EACf;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;;UAGe;EACf,QAAQ;EACR;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX;;UAGe;EACf,MAAM;EACN,aAAa;EACb,SAAS;EACT,cAAc;;UAGC;EACf;EACA;EACA;EACA;EACA;EACA;EACA,WAAW;EACX;EACA;EACA;EACA;EACA;EACA;;;EAGA,iBAAiB;;;EAGjB,YAAY;;;;KAKF;EACN;EAAe;EAAoB;;EACnC;EAAmB;;;;iBCvFT,oBAAoB,gBAAgB;;;;;;;;;;iBAuB7B,yBAAyB,eAAe,eAAe;;;;;iBA6BxD,wBAAwB,eAAe,QAAQ;;;iBCDrD,iBACd,SAAS,gCACR;iBAIa,sBACd,SAAS,gCACR;iBAIa,oBACd,SAAS,gCACR;iBAIa,oBACd,SAAS,gCACR;;iBAKa,cACd,SAAS,gCACR;;;;UCrDc;;;;;WAKN;;WAEA,QAAQ;;;;;WAKR;;;;;WAKA;aACE;;aAEA;;;;KAKD;;UAUK;WACN;WACA;WACA;WACA;WACA,aAAa;;;UAIP;WACN;WACA,UAAU,SAAS,QAAQ,OAAO;;WAElC,QAAQ;aACN;aACA,QAAQ;;;;;;;;;KAUT;WAEG;WACA,OAAO;WACP;WACA,mBAAmB;;WAEnB;;WAEA;;WAEA;WACA,WAAW,SAAS,QAAQ,OAAO;WACnC,YAAY;;WAGZ;WACA,OAAO;WACP;;WAEA;WACA,YAAY;;;;UCvGV;;;EAGf;;EAEA;;;EAGA;;;EAGA,WAAW;;UAGI;EACf;;;EAGA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;EAEA;;;EAGA,WAAW;;;EAGX,SAAS;;UAGM;;EAEf,SAAS;;;EAGT,OAAO;;;;EAIP;IAAU;IAAa;;;;;EAIvB;;UAGe;EACf,MAAM;;;EAGN,aAAa;IAAQ;IAAe;IAAe;;;iBAGrC,kBAAkB,MAAM,2BAA2B;;;UC3BlD;EACf,OAAO;;EAEP,eAAe;;EAEf;;;;;;EAMA,eAAe,eAAe,gBAAgB;;iBAGhC,cAAc,MAAM,uBAAuB"}

package/dist/index.d.ts

@@ -3,46 +3,46 @@ import { a as JudgeError, c as ValidationError, i as ConfigError, n as AgentEval
 import { _ as toAgentProfileJson, g as groupRunsByAgentProfileCell, h as buildAgentProfileCell, i as AgentProfileCellInput, l as AgentProfileJson, m as agentProfileCellKey, p as AgentProfileSourceInput, r as AgentProfileCell, s as AgentProfileDimensionValue, t as AGENT_PROFILE_KINDS, y as verifyAgentProfileCell } from "./agent-profile-cell-BcpFbNd5.js";
 import { E as isToolSpan, S as TraceEvent, T as isLlmSpan, _ as Span, a as FAILURE_CLASSES, c as JudgeSpan, f as Run, h as RunStatus, l as LlmSpan, n as BudgetLedgerEntry, o as FailureClass, p as RunCaptureReport, r as BudgetSpec, s as GenericSpan, t as Artifact, w as isJudgeSpan, x as ToolSpan } from "./schema-D6EVKzTB.js";
 import { _ as runTaskScore, a as RunRecord, c as RunSplitTag, d as RunTokenUsage, f as RunTraceRef, g as parseRunRecordSafe, h as modelHasSnapshot, i as RunOutcome, l as RunTaskFailure, m as isRunRecord, n as RunCostProvenance, o as RunRecordValidationError, p as UNKNOWN_MODEL, r as RunJudgeMetadata, s as RunSearchCoordinates, t as JudgeScoresRecord, u as RunTerminalOutcome, v as validateRunRecord } from "./run-record-C3awiXPf.js";
-import { $ as QueryTracesPage, H as createBoundedTraceAnalysisStore, J as TraceAnalysisStore, Q as ErrorCluster, S as makeProposalFinding, X as DEFAULT_TRACE_ANALYST_BUDGETS, Z as DatasetOverview, at as TraceAnalystSpan, b as computeFindingId, c as AnalystRunEvent, ct as TraceAnalystTraceSummary, d as AnalystRunSummary, dt as ViewTraceResult, et as SearchSpanResult, f as AnalystSeverity, i as AnalystFinding, it as TraceAnalystFilters, l as AnalystRunInputs, lt as ViewSpansResult, m as EvidenceRef, n as AnalystContext, nt as SpanMatchRecord, ot as TraceAnalystSpanKind, p as AnalystUsageReceipt, rt as TraceAnalystByteBudgets, st as TraceAnalystSpanStatus, t as Analyst, tt as SearchTraceResult, u as AnalystRunResult, ut as ViewTraceOversized, v as ProposalFinding, x as makeFinding } from "./types-DGImvHIr.js";
-import { $ as AssertServedModelOptions, A as createChatClient, B as LlmToolDefinition, C as ChatResponse, D as MockTransportOpts, E as CustomTransportOpts, F as LlmMessage, H as FileSystemRawProviderSink, K as NoopRawProviderSink, L as LlmTokenLogprob, M as LlmCallRequest, N as LlmCallResult, O as OpenAiCompatibleTransportOpts, P as LlmChargeBounds, Q as AssertCrossFamilyServedOptions, R as LlmToolCall, S as ChatRequest, T as CreateChatClientOpts, V as LlmUsage, W as InMemoryRawProviderSink, X as RawProviderSink, Y as RawProviderEvent, _ as ScenarioResult, a as CompletionCriterion, at as assertServedModel, b as ChatCallOpts, c as JudgeFn, ct as AssertCrossFamilyOptions, d as PersonaConfig, dt as assertCrossFamily, et as ModelSubstitutionError, f as ProductClientConfig, ft as judgeFamily, g as ScenarioFile, h as RubricDimension, i as CollectedArtifacts, it as assertCrossFamilyServed, j as LlmCallMetadata, k as SandboxSdkTransportOpts, l as JudgeInput, lt as CrossFamilyError, m as RouteMap, n as ArtifactResult, nt as ServedModelPolicy, o as DimensionJudgeScore, ot as assertServedModels, p as ProductScenario, r as CheckResult, rt as ServedModelVerdict, s as DriverState, st as checkServedModel, t as ArtifactCheck, tt as ServedModelCheck, u as JudgeRubric, ut as JudgeFamily, v as Turn, w as ChatTransport, x as ChatClient, y as TurnResult, z as LlmToolChoice } from "./types-eMhdacVp.js";
+import { $ as QueryTracesPage, H as createBoundedTraceAnalysisStore, J as TraceAnalysisStore, Q as ErrorCluster, S as makeProposalFinding, X as DEFAULT_TRACE_ANALYST_BUDGETS, Z as DatasetOverview, at as TraceAnalystSpan, b as computeFindingId, c as AnalystRunEvent, ct as TraceAnalystTraceSummary, d as AnalystRunSummary, dt as ViewTraceResult, et as SearchSpanResult, f as AnalystSeverity, i as AnalystFinding, it as TraceAnalystFilters, l as AnalystRunInputs, lt as ViewSpansResult, m as EvidenceRef, n as AnalystContext, nt as SpanMatchRecord, ot as TraceAnalystSpanKind, p as AnalystUsageReceipt, rt as TraceAnalystByteBudgets, st as TraceAnalystSpanStatus, t as Analyst, tt as SearchTraceResult, u as AnalystRunResult, ut as ViewTraceOversized, v as ProposalFinding, x as makeFinding } from "./types-6Lfdlshp.js";
+import { $ as AssertCrossFamilyServedOptions, A as createChatClient, B as LlmToolChoice, C as ChatResponse, D as MockTransportOpts, E as CustomTransportOpts, G as InMemoryRawProviderSink, H as LlmUsage, I as LlmMessage, M as LlmCallRequest, N as LlmCallResult, O as OpenAiCompatibleTransportOpts, P as LlmChargeBounds, R as LlmTokenLogprob, S as ChatRequest, T as CreateChatClientOpts, U as FileSystemRawProviderSink, V as LlmToolDefinition, X as RawProviderEvent, Z as RawProviderSink, _ as ScenarioResult, a as CompletionCriterion, at as assertCrossFamilyServed, b as ChatCallOpts, c as JudgeFn, ct as checkServedModel, d as PersonaConfig, dt as CrossFamilyError, et as AssertServedModelOptions, f as ProductClientConfig, ft as JudgeFamily, g as ScenarioFile, h as RubricDimension, i as CollectedArtifacts, it as ServedModelVerdict, j as LlmCallMetadata, k as SandboxSdkTransportOpts, l as JudgeInput, lt as describeServedModelCheck, m as RouteMap, mt as judgeFamily, n as ArtifactResult, nt as ServedModelCheck, o as DimensionJudgeScore, ot as assertServedModel, p as ProductScenario, pt as assertCrossFamily, q as NoopRawProviderSink, r as CheckResult, rt as ServedModelPolicy, s as DriverState, st as assertServedModels, t as ArtifactCheck, tt as ModelSubstitutionError, u as JudgeRubric, ut as AssertCrossFamilyOptions, v as Turn, w as ChatTransport, x as ChatClient, y as TurnResult, z as LlmToolCall } from "./types-B_IFcKp5.js";
 import { a as ContinuousCalibrationResult, d as calibrateJudgeContinuous, f as continuousAgreement, h as verbosityBias, i as ContinuousAgreementOptions, l as VerbosityBiasResult, n as CandidateScore, o as GoldenItem, r as ContinuousAgreement, t as CalibrationResult, u as calibrateJudge } from "./judge-calibration-DFtEMlde.js";
-import { A as JudgeGateInput, B as measureJudgeAgreement, C as CalibrationMerge, D as JudgeConfusion, E as JudgeCalibrationSet, F as JudgeRunMeta, Gt as CorpusAgreementOptions, H as snapshotFromJudgeAgreement, I as OwnerVerdict, Jt as CorpusScoreRecord, Kt as CorpusAgreementPerDimension, L as addCalibrationVerdicts, M as JudgeGateRefusedError, N as JudgeGateStatus, O as JudgeExampleAgreement, P as JudgeRun, R as assertJudgeMayGate, S as CalibrationExample, T as JudgeAgreement, V as registerCalibrationSet, Xt as corpusInterRaterAgreementFromJudgeScores, Yt as corpusInterRaterAgreement, Zt as interRaterReliability, ct as SeriesConvergenceResult, j as JudgeGatePolicy, k as JudgeGateDecision, lt as analyzeSeries, qt as CorpusAgreementReport, st as SeriesConvergenceOptions, w as DEFAULT_JUDGE_GATE_POLICY, z as judgeGateDecision } from "./index-DWd1gUwR.js";
+import { $ as measureJudgeAgreement, B as JudgeConfusion, F as CalibrationExample, G as JudgeGateRefusedError, H as JudgeGateDecision, I as CalibrationMerge, J as JudgeRunMeta, K as JudgeGateStatus, L as DEFAULT_JUDGE_GATE_POLICY, Q as judgeGateDecision, R as JudgeAgreement, St as analyzeSeries, U as JudgeGateInput, V as JudgeExampleAgreement, W as JudgeGatePolicy, X as addCalibrationVerdicts, Y as OwnerVerdict, Z as assertJudgeMayGate, an as CorpusAgreementPerDimension, bt as SeriesConvergenceOptions, cn as corpusInterRaterAgreement, et as registerCalibrationSet, in as CorpusAgreementOptions, ln as corpusInterRaterAgreementFromJudgeScores, on as CorpusAgreementReport, q as JudgeRun, sn as CorpusScoreRecord, tt as snapshotFromJudgeAgreement, un as interRaterReliability, xt as SeriesConvergenceResult, z as JudgeCalibrationSet } from "./index-CfJULcW5.js";
 import { a as partialCredit, c as spearmanR, d as weightedMean, i as confidenceInterval, l as summarizeNumberSeries, n as WeightedCompositeInput, o as pearsonR, r as WeightedCompositeResult, s as ranks, t as SeriesDistribution, u as weightedComposite } from "./descriptive-B2iPaT9J.js";
 import { n as bonferroni, r as holm, t as benjaminiHochberg } from "./multiplicity-DIWHvysC.js";
-import { A as PairRunRecordsResult, At as EProcess, B as ArmPopulationSummary, Bt as mcnemar, D as MatchedRunRecordPair, E as MatchedPair, F as comparePairedArms, Ft as ExactRiskDifferenceResult, Gt as wilson, Ht as pairedRiskDifferenceExact, I as pairArms, It as McNemarResult, L as pairRunRecords, Lt as ProportionInterval, M as PairedArmsComparison, Mt as EProcessState, N as PairedCorrectness, Nt as EProcessStep, O as PairArmsOptions, P as PairedMetricDelta, Pt as eProcess, R as ArmAssignment, Rt as RiskDifferenceResult, S as manifestContentDigest, T as ComparePairedArmsOptions, Ut as pairedRiskDifferenceScore, Vt as pairedRiskDifference, Wt as passAtK, _t as AttestationProvenance, bt as attest, gt as ATTESTATION_ALGORITHM, it as paretoFrontier, j as PairedArmRow, jt as EProcessOptions, k as PairArmsResult, nt as ParetoResult, rt as dominates, tt as Objective, vt as AttestationVerification, x as hashJson, xt as verifyAttestation, yt as AttestedReport, z as ArmPopulationReading, zt as ScoreRiskDifferenceResult } from "./statistical-heldout-pD1Yah-0.js";
+import { A as PairRunRecordsResult, At as EProcess, B as ArmPopulationSummary, Bt as mcnemar, D as MatchedRunRecordPair, E as MatchedPair, F as comparePairedArms, Ft as ExactRiskDifferenceResult, Gt as wilson, Ht as pairedRiskDifferenceExact, I as pairArms, It as McNemarResult, L as pairRunRecords, Lt as ProportionInterval, M as PairedArmsComparison, Mt as EProcessState, N as PairedCorrectness, Nt as EProcessStep, O as PairArmsOptions, P as PairedMetricDelta, Pt as eProcess, R as ArmAssignment, Rt as RiskDifferenceResult, S as manifestContentDigest, T as ComparePairedArmsOptions, Ut as pairedRiskDifferenceScore, Vt as pairedRiskDifference, Wt as passAtK, _t as AttestationProvenance, bt as attest, gt as ATTESTATION_ALGORITHM, it as paretoFrontier, j as PairedArmRow, jt as EProcessOptions, k as PairArmsResult, nt as ParetoResult, rt as dominates, tt as Objective, vt as AttestationVerification, x as hashJson, xt as verifyAttestation, yt as AttestedReport, z as ArmPopulationReading, zt as ScoreRiskDifferenceResult } from "./statistical-heldout-BjzSOmLH.js";
 import { a as PairedPromotionDecision, c as BOOTSTRAP_GATE_MIN_N, d as PairedSignTestResult, f as PairedTTestResult, g as pairedTTest, h as pairedSignTest, i as PairedMcNemarEvidence, l as PairedBootstrapOptions, m as pairedBootstrap, n as PairedDecisionShape, o as PairedPromotionDecisionOptions, p as SignTestAlternative, r as PairedDecisionStatistic, s as decidePairedPromotion, t as PairedDecisionMethod, u as PairedBootstrapResult } from "./paired-promotion-decision-CQMJ5xGN.js";
 import { a as requiredPairedSampleSize, i as pairedMde, n as mcnemarPower, o as requiredSampleSize, r as mcnemarRequiredN, t as mulberry32 } from "./random-DGEyr3zG.js";
 import { C as mannWhitneyU, S as WilcoxonSignedRankResult, _ as evaluateReleaseConfidence, a as bootstrapCi, b as RankTestMethodRequest, d as ReleaseConfidenceIssue, f as ReleaseConfidenceMetrics, g as ReleaseTraceEvidence, i as Verdict, n as BootstrapResult, p as ReleaseConfidenceScorecard, s as ActionableSideInfo, t as BootstrapOptions, u as ReleaseConfidenceInput, v as MannWhitneyResult, w as wilcoxonSignedRank, x as RankTestOptions, y as RankTestMethod } from "./promotion-gate-CkVZHagC.js";
-import { $ as summarizeBackendIntegrity, Bi as runCampaign, Ct as RequirementCheck, Dt as createLlmCorrectnessChecker, Et as completionVerdict, Li as CampaignCellFailureReceipt, Ot as verifyCompletion, Q as assertRealBackend, Ri as CampaignCellRetryPolicy, Sr as runCanaries, St as ProducedState, Tt as TaskGold, X as BackendIntegrityError, Z as BackendIntegrityReport, _r as CanaryAlert, _t as CompletionRequirement, bi as LlmJudgeOptions, br as CanaryOptions, bt as LlmCorrectnessCheckerOpts, dr as RedTeamCategory, fr as RedTeamFinding, ft as ArtifactEventLike, gr as scoreRedTeamOutput, gt as extractProducedState, hr as redTeamReport, ht as ToolCallEventLike, lr as DEFAULT_RED_TEAM_CORPUS, mr as redTeamDataset, mt as RuntimeEventLike, pr as RedTeamReport, pt as ProposalEventLike, ur as RedTeamCase, vr as CanaryEvaluation, vt as CompletionVerdict, wi as transientDispatchFailure, wt as SatisfiedBy, xi as llmJudge, xr as CanaryReport, xt as ProducedProposal, yr as CanaryKind, yt as CorrectnessChecker, zi as RunCampaignOptions } from "./index-BvX3CFNS.js";
-import { R as Scenario, S as JudgeConfig, _ as GateDecision, a as CampaignResult, w as JudgeScore } from "./types-DDKfH-N6.js";
+import { $ as summarizeBackendIntegrity, Bi as runCampaign, Ct as RequirementCheck, Dt as createLlmCorrectnessChecker, Et as completionVerdict, Li as CampaignCellFailureReceipt, Ot as verifyCompletion, Q as assertRealBackend, Ri as CampaignCellRetryPolicy, Sr as runCanaries, St as ProducedState, Tt as TaskGold, X as BackendIntegrityError, Z as BackendIntegrityReport, _r as CanaryAlert, _t as CompletionRequirement, bi as LlmJudgeOptions, br as CanaryOptions, bt as LlmCorrectnessCheckerOpts, dr as RedTeamCategory, fr as RedTeamFinding, ft as ArtifactEventLike, gr as scoreRedTeamOutput, gt as extractProducedState, hr as redTeamReport, ht as ToolCallEventLike, lr as DEFAULT_RED_TEAM_CORPUS, mr as redTeamDataset, mt as RuntimeEventLike, pr as RedTeamReport, pt as ProposalEventLike, ur as RedTeamCase, vr as CanaryEvaluation, vt as CompletionVerdict, wi as transientDispatchFailure, wt as SatisfiedBy, xi as llmJudge, xr as CanaryReport, xt as ProducedProposal, yr as CanaryKind, yt as CorrectnessChecker, zi as RunCampaignOptions } from "./index-CdEVr8-O.js";
+import { R as Scenario, S as JudgeConfig, _ as GateDecision, a as CampaignResult, w as JudgeScore } from "./types-M6eF9m-r.js";
 import { a as RunFilter, i as InMemoryTraceStore, n as FileSystemTraceStore, o as SpanFilter, s as TraceStore, t as EventFilter } from "./store-B3nWRy3B.js";
 import { a as MintedRolloutLine } from "./schema-C-ziyJD7.js";
 import { n as ScorePreference } from "./reward-BeUgVVC_.js";
 import { a as mintRolloutRows, i as RolloutSearchLineage, n as MintRolloutResult, t as MintRolloutOptions } from "./mint-CkzpVPdz.js";
-import { a as ProfileAxisSpec, c as expandProfileAxes, i as HarnessType, l as harnessAxisOf, n as CODING_HARNESSES, o as agentProfileHash, r as HARNESS_NATIVE_MODEL, s as agentProfileId, t as AgentProfile } from "./agent-profile-Bc7ADJY4.js";
+import { a as ProfileAxisSpec, c as expandProfileAxes, i as HarnessType, l as harnessAxisOf, n as CODING_HARNESSES, o as agentProfileHash, r as HARNESS_NATIVE_MODEL, s as agentProfileId, t as AgentProfile } from "./agent-profile-CNmwfaqq.js";
 import { i as DatasetSplit, n as DatasetManifest, r as DatasetScenario, t as Dataset } from "./dataset-DQqhOCPt.js";
 import { C as runEquivalenceCheck, S as defineEquivalenceCheck, _ as StrategyChecker, a as CheckerIdentity, b as VerificationStrategyProfile, c as EquivalenceCheckDefinition, d as EquivalenceCheckerInput, f as EquivalenceCheckerResult, g as EquivalenceRecord, h as EquivalenceProtocolError, i as equivalenceVerdict, l as EquivalenceCheckSpec, m as EquivalenceObligationStatus, n as VerdictCertification, o as CheckerOutcome, p as EquivalenceObligation, r as certificationEvidenceDigest, s as EquivalenceArm, t as DefaultVerdict, u as EquivalenceChecker, v as VERIFICATION_STRATEGIES, x as VerificationStrategySource, y as VERIFICATION_STRATEGY_SOURCES } from "./verdict-C24NELtN.js";
 import { a as MultiLayerVerifier, c as VerifyOptions, i as LayerStatus, l as gradeSemanticStatus, n as Layer, o as Severity, r as LayerResult, s as VerificationReport, t as Finding } from "./multi-layer-verifier-DEgMUk6v.js";
 import { C as HeldOutGateRejectionCode, S as HeldOutGateDecision, _ as gainHistogram, a as ResearchReportOptions, b as HeldOutGate, d as GainDistributionOptions, g as SummaryTableRow, h as SummaryTableOptions, l as GainDistributionBin, m as SummaryTable, p as ParetoPoint, t as ResearchReport, u as GainDistributionFigureSpec, v as summaryTable, w as SplitCoverage, x as HeldOutGateConfig, y as GateEvidence } from "./summary-report-C8V49Jt3.js";
 import { a as FailureClassification, c as FailureRule, d as failureBlame, i as FailureBlame, l as classifyFailure, o as FailureContext, r as failureClusterView, s as FailureReasonRule, u as classifyFailureReason } from "./failure-cluster-DN0t63Z0.js";
-import { d as InsightReport, n as analyzeRuns, t as AnalyzeRunsOptions } from "./analyze-runs-D7In4c3z.js";
-import { _ as SelfImproveMethodResult, a as DefinedAgentEval, c as SelfImproveMethodOptions, d as SelfImproveProposerOptions, f as SelfImproveProposerResult, g as SelfImproveMethodProvenance, h as selfImprove, i as DefineAgentEvalOptions, l as SelfImproveOptions, n as AgentEvalEvaluateOptions, o as defineAgentEval, p as SelfImproveResult, r as AgentEvalImproveOptions, s as SelfImproveBudget, t as AgentEvalAgent, u as SelfImproveProgressEvent } from "./define-agent-eval-Bsoi2GUV.js";
-import { B as FindingsDiff, H as PersistedFinding, U as diffFindings, V as FindingsStore, z as DiffPolicy } from "./systemone-D3MrW63C.js";
-import { i as ExactAnalystRunEvent, l as ExactCapableAnalyst, o as ExactAnalystRunResult } from "./exact-types-CVn1QxFO.js";
-import { a as createTraceAnalyst, i as TraceAnalystDefinition, n as buildDefaultAnalystRegistry, r as CreateTraceAnalystOptions, t as DefaultAnalystRegistryOptions } from "./default-registry-BMHqNR6_.js";
-import { r as TraceAnalysisEngineResult, s as RawAnalystFinding, t as TraceAnalysisEngine } from "./engine-BF95CWzH.js";
-import { a as BudgetPolicy, c as RegistryRunOpts, i as AnalystRegistryOptions, s as ExactRegistryRunOpts, t as AnalystRegistry } from "./registry-DDKaUrmM.js";
-import { C as SEMANTIC_CONCEPT_JUDGE_VERSION, D as runSemanticConceptJudge, E as SemanticConceptJudgeResult, S as ConceptSpec, T as SemanticConceptJudgeOptions, _ as DspyRlmTraceEngineOptions, a as FAILURE_MODE_KIND_SPEC, b as createChatTraceEngine, d as FindingSubject, f as FindingSubjectKind, t as DEFAULT_TRACE_ANALYST_KINDS, v as createDspyRlmTraceEngine, w as SemanticConceptJudgeInput, x as ConceptFinding, y as ChatTraceEngineOptions } from "./index-BJZ7s-sv.js";
-import { i as BenchmarkEvaluation } from "./types-IERw_HxR.js";
+import { d as InsightReport, n as analyzeRuns, t as AnalyzeRunsOptions } from "./analyze-runs-DfX6dHP0.js";
+import { _ as SelfImproveMethodResult, a as DefinedAgentEval, c as SelfImproveMethodOptions, d as SelfImproveProposerOptions, f as SelfImproveProposerResult, g as SelfImproveMethodProvenance, h as selfImprove, i as DefineAgentEvalOptions, l as SelfImproveOptions, n as AgentEvalEvaluateOptions, o as defineAgentEval, p as SelfImproveResult, r as AgentEvalImproveOptions, s as SelfImproveBudget, t as AgentEvalAgent, u as SelfImproveProgressEvent } from "./define-agent-eval-JKt3gMQr.js";
+import { B as FindingsDiff, H as PersistedFinding, U as diffFindings, V as FindingsStore, z as DiffPolicy } from "./systemone-CHnqwmYg.js";
+import { i as ExactAnalystRunEvent, l as ExactCapableAnalyst, o as ExactAnalystRunResult } from "./exact-types-Do63KoKn.js";
+import { a as createTraceAnalyst, i as TraceAnalystDefinition, n as buildDefaultAnalystRegistry, r as CreateTraceAnalystOptions, t as DefaultAnalystRegistryOptions } from "./default-registry-CLG5YqGS.js";
+import { r as TraceAnalysisEngineResult, s as RawAnalystFinding, t as TraceAnalysisEngine } from "./engine-dtRVeuQs.js";
+import { a as BudgetPolicy, c as RegistryRunOpts, i as AnalystRegistryOptions, s as ExactRegistryRunOpts, t as AnalystRegistry } from "./registry-BcdJmyoU.js";
+import { C as SEMANTIC_CONCEPT_JUDGE_VERSION, D as runSemanticConceptJudge, E as SemanticConceptJudgeResult, S as ConceptSpec, T as SemanticConceptJudgeOptions, _ as DspyRlmTraceEngineOptions, a as FAILURE_MODE_KIND_SPEC, b as createChatTraceEngine, d as FindingSubject, f as FindingSubjectKind, t as DEFAULT_TRACE_ANALYST_KINDS, v as createDspyRlmTraceEngine, w as SemanticConceptJudgeInput, x as ConceptFinding, y as ChatTraceEngineOptions } from "./index-CTIRz-AB.js";
+import { i as BenchmarkEvaluation } from "./types-DZ0mYVw9.js";
 import { a as SpanHandle, n as CaptureWrite, o as TraceEmitter, s as TraceEmitterOptions, t as CaptureErrorHandler } from "./emitter-DABYYTcH.js";
 import { n as SandboxDriver } from "./sandbox-harness-BlSOu4LX.js";
 import { _ as shareAllowed, a as RedactionProfile, d as assessShareSafety, g as redactText, h as redactForShare, l as ShareSafetyStatus, m as redact, n as REDACTION_VERSION, o as RedactionReport, r as RedactOptions, u as ShareSafetyVerdict } from "./redact-DXdlFb_U.js";
 import { a as PairedEvalueStep, c as SequentialDecision, i as PairedEvalueSequence, l as evaluateInterimReleaseConfidence, n as InterimReleaseConfidenceInput, r as PairedEvalueOptions, t as InterimReleaseConfidence, u as pairedEvalueSequence } from "./sequential-KwGspfeR.js";
 import { S as welchsTTest, _ as computeToolUseMetrics, b as WelchTestStatus, d as Trajectory, f as TrajectoryStep, g as ToolUseOptions, h as ToolUseMetrics, i as toolWasteView, m as ToolStats, p as buildTrajectory, s as judgeAgreementView, u as budgetBreachView, x as iqr, y as WelchTestResult } from "./tool-waste-CkjzWmeY.js";
-import { A as AnalystFindingDigest, B as ControlRuntimeConfig, C as createFeedbackTrajectory, D as renderPreferenceMemoryMarkdown, E as feedbackTrajectoryToOptimizerRow, F as ControlBudget, H as ControlSeverity, I as ControlContext, L as ControlDecision, M as AnalystRunDigest, N as analystRunDigest, O as summarizePreferenceMemory, P as ControlActionOutcome, R as ControlEvalResult, S as controlRunToFeedbackTrajectory, T as feedbackTrajectoriesToOptimizerRows, U as ControlStep, V as ControlRuntimeError, W as StopDecision, _ as PreferenceMemoryEntry, a as FeedbackLabel, b as analystRunToReviewRequests, c as FeedbackOptimizerRow, d as FeedbackTask, f as FeedbackTrajectory, g as InMemoryFeedbackTrajectoryStore, h as FileSystemFeedbackTrajectoryStore, i as FeedbackAttempt, j as AnalystReviewDecision, k as withAssignedFeedbackSplit, l as FeedbackOutcome, m as FeedbackTrajectoryStore, n as AnalystReviewRequest, o as FeedbackLabelKind, p as FeedbackTrajectoryFilter, r as FeedbackArtifactType, s as FeedbackLabelSource, t as AnalystFeedbackTrajectoryOptions, u as FeedbackSplitPolicy, v as ProposedSideEffect, w as feedbackTrajectoriesToDatasetScenarios, x as assignFeedbackSplit, y as analystRunToFeedbackTrajectory, z as ControlRunResult } from "./feedback-trajectory-BZhVCWuz.js";
-import { A as TraceInsightReadiness, F as describeTraceInsightScope, G as AnalyzeTracesResult, H as tokenizeDomainWords, I as domainEvidencePattern, K as analyzeTraces, L as inferDomainKeywords, N as buildTraceInsightContext, P as buildTraceInsightPrompt, V as scoreTraceInsightReadiness, a as captureFetchToRawSink, b as OtlpSpan, c as ExtractedUsage, d as extractUsageFromSse, f as OtlpFileTraceStore, g as spanRecordsToTraceAnalysisStore, h as otlpTextToTraceAnalysisStore, j as TraceInsightSuite, m as SpanRecordsToTraceAnalysisStoreOptions, n as toolSpansToTraceAnalysisStore, o as OtlpFlatLine, u as extractUsage, v as OtlpExport, x as exportRunAsOtlp } from "./store-tool-spans-Bf9OVKs6.js";
-import { a as RunIntegrityReport, o as assertRunCaptured, t as RunIntegrityError } from "./integrity-CCwvezWS.js";
-import { a as SteeringChange, c as CampaignRunContext, d as CampaignVariant, f as EvalCampaignOptions, h as runEvalCampaign, i as Researcher, l as CampaignRunOutcome, m as FailedRun, n as ExperimentResult, o as CampaignFactoryParams, p as EvalCampaignResult, r as FailureMode, s as CampaignIntegrityPolicy, t as ExperimentPlan, u as CampaignScenario } from "./researcher-H0p44OYC.js";
+import { A as AnalystFindingDigest, B as ControlRuntimeConfig, C as createFeedbackTrajectory, D as renderPreferenceMemoryMarkdown, E as feedbackTrajectoryToOptimizerRow, F as ControlBudget, H as ControlSeverity, I as ControlContext, L as ControlDecision, M as AnalystRunDigest, N as analystRunDigest, O as summarizePreferenceMemory, P as ControlActionOutcome, R as ControlEvalResult, S as controlRunToFeedbackTrajectory, T as feedbackTrajectoriesToOptimizerRows, U as ControlStep, V as ControlRuntimeError, W as StopDecision, _ as PreferenceMemoryEntry, a as FeedbackLabel, b as analystRunToReviewRequests, c as FeedbackOptimizerRow, d as FeedbackTask, f as FeedbackTrajectory, g as InMemoryFeedbackTrajectoryStore, h as FileSystemFeedbackTrajectoryStore, i as FeedbackAttempt, j as AnalystReviewDecision, k as withAssignedFeedbackSplit, l as FeedbackOutcome, m as FeedbackTrajectoryStore, n as AnalystReviewRequest, o as FeedbackLabelKind, p as FeedbackTrajectoryFilter, r as FeedbackArtifactType, s as FeedbackLabelSource, t as AnalystFeedbackTrajectoryOptions, u as FeedbackSplitPolicy, v as ProposedSideEffect, w as feedbackTrajectoriesToDatasetScenarios, x as assignFeedbackSplit, y as analystRunToFeedbackTrajectory, z as ControlRunResult } from "./feedback-trajectory-CYVJXolG.js";
+import { A as TraceInsightReadiness, F as describeTraceInsightScope, G as AnalyzeTracesResult, H as tokenizeDomainWords, I as domainEvidencePattern, K as analyzeTraces, L as inferDomainKeywords, N as buildTraceInsightContext, P as buildTraceInsightPrompt, V as scoreTraceInsightReadiness, a as captureFetchToRawSink, b as OtlpSpan, c as ExtractedUsage, d as extractUsageFromSse, f as OtlpFileTraceStore, g as spanRecordsToTraceAnalysisStore, h as otlpTextToTraceAnalysisStore, j as TraceInsightSuite, m as SpanRecordsToTraceAnalysisStoreOptions, n as toolSpansToTraceAnalysisStore, o as OtlpFlatLine, u as extractUsage, v as OtlpExport, x as exportRunAsOtlp } from "./store-tool-spans-DxsFwtdr.js";
+import { a as RunIntegrityReport, o as assertRunCaptured, t as RunIntegrityError } from "./integrity-D9lBnv_K.js";
+import { a as SteeringChange, c as CampaignRunContext, d as CampaignVariant, f as EvalCampaignOptions, h as runEvalCampaign, i as Researcher, l as CampaignRunOutcome, m as FailedRun, n as ExperimentResult, o as CampaignFactoryParams, p as EvalCampaignResult, r as FailureMode, s as CampaignIntegrityPolicy, t as ExperimentPlan, u as CampaignScenario } from "./researcher-DKWtoMWk.js";
 import "./index-D7fmC-1K.js";
 import { y as OUTPUT_VALUE } from "./attribute-vocabulary-CcXqHODT.js";
 import { i as judgeSpans, n as argHash, s as runsForScenario } from "./query-Cmv_U8vf.js";
@@ -3836,5 +3836,5 @@ export declare function analyzeAntiSlop(outputs: string[], config: Omit<Required
   penaltyWeights: Record<SlopCategory, number>;
 }): AntiSlopReport;
 //#endregion
-export { AGENT_PROFILE_KINDS, ATTESTATION_ALGORITHM, type ActionExecutionPolicy, type ActionPolicyDecision, type ActionableSideInfo, type ActiveLearningOptions, type AgentEvalAgent, AgentEvalError, type AgentEvalErrorCode, type AgentEvalEvaluateOptions, type AgentEvalImproveOptions, type AgentProfile, type AgentProfileCell, type AgentProfileCellInput, type AgentProfileDimensionValue, type AgentProfileJson, type AgentProfileSourceInput, type Analyst, type AnalystContext, type AnalystFeedbackTrajectoryOptions, type AnalystFinding, type AnalystFindingDigest, AnalystRegistry, type AnalystRegistryOptions, type AnalystReviewDecision, type AnalystReviewRequest, type AnalystRunDigest, type AnalystRunEvent, type AnalystRunInputs, type AnalystRunResult, type AnalystRunSummary, type AnalystSeverity, type AnalystUsageReceipt, type AnalyzeRunsOptions, type AnalyzeTracesResult, type ArgumentCheck, type ArmAssignment, type ArmPopulationReading, type ArmPopulationSummary, type Artifact, type ArtifactCheck, type ArtifactEventLike, type ArtifactResult, type AssertCapabilityHeadroomOptions, type AssertCrossFamilyOptions, type AssertCrossFamilyServedOptions, type AssertServedModelOptions, type AssertSingleBackendOptions, type AttestationProvenance, type AttestationVerification, type AttestedReport, BOOTSTRAP_GATE_MIN_N, type BackendDescriptor, BackendIntegrityError, type BackendIntegrityReport, type BenchmarkEvaluation, type BlendWeights, type BootstrapOptions, type BootstrapResult, type BoundedProcessInput, type BoundedProcessResult, type BudgetLedgerEntry, type BudgetPolicy, type BudgetSpec, CODING_HARNESSES, type CalibrationExample, type CalibrationMerge, type CalibrationResult, type CampaignCellFailureReceipt, type CampaignCellRetryPolicy, type CampaignFactoryParams, type CampaignIntegrityPolicy, type CampaignResult, type CampaignRunContext, type CampaignRunOutcome, type CampaignScenario, type CampaignVariant, type CanaryAlert, type CanaryEvaluation, type CanaryKind, type CanaryLeak, type CanaryOptions, type CanaryReport, type CandidateScore, type CapabilityHeadroomOptions, type CapabilityHeadroomResult, type CaptureErrorHandler, type CaptureWrite, type CellVerdict, type ChannelRollup, type ChatCallOpts, type ChatClient, type ChatRequest, type ChatResponse, type ChatTraceEngineOptions, type ChatTransport, type CheckResult, type CheckerIdentity, type CheckerOutcome, type CliffsMagnitude, type CollectedArtifacts, type CommandRunner, type ComparePairedArmsOptions, type CompletionCriterion, type CompletionRequirement, type CompletionVerdict, type ConceptFinding, type ConceptSpec, ConfigError, type ContinuousAgreement, type ContinuousAgreementOptions, type ContinuousCalibrationResult, type ContractCheckResult, type ContractLintFinding, type ContractRule, type ContractRuleExecution, type ContractRun, type ContractSpan, type ContractStatus, type ContractVerdict, type ControlActionOutcome, type ControlBudget, type ControlContext, type ControlDecision, type ControlEvalResult, type ControlRunResult, type ControlRuntimeConfig, type ControlRuntimeError, type ControlStep, type CorpusAgreementOptions, type CorpusAgreementPerDimension, type CorpusAgreementReport, type CorpusScoreRecord, type CorrectnessChecker, type CostChannel, type CostEntry, CostLedger, type CostLedgerFilter, type CostLedgerHandle, type CostLedgerOptions, type CostLedgerPersistence, type CostLedgerSummary, type CostProvenance, type CostReceipt, type CostReceiptInput, type CostResult, type CostSummary, type CostUsage, type CounterfactualContext, type CounterfactualMutation, type CounterfactualResult, type CounterfactualRunner, type CreateChatClientOpts, type CreateTraceAnalystOptions, CrossFamilyError, type CustomTokenPricing, type CustomTransportOpts, DEFAULT_JUDGE_GATE_POLICY, DEFAULT_RED_TEAM_CORPUS, DEFAULT_TRACE_ANALYST_BUDGETS, DEFAULT_TRACE_ANALYST_KINDS, type DataAcquisitionPlan, type DatasetManifest, type DatasetOverview, type DatasetScenario, type DatasetSplit, type DefaultAnalystRegistryOptions, type DefaultVerdict, type DefineAgentEvalOptions, type DefinedAgentEval, type DeployGateLayerInput, type DeployRunResult, type DeployRunner, type DetectorEvent, type DetectorSignal, type DiffPolicy, type DiffScorecardOptions, type DimensionJudgeScore, type DirEntry, type DiscoverPersonasOptions, type DiscoveredPersona, type DriverState, type DspyRlmTraceEngineOptions, type EProcess, type EProcessOptions, type EProcessState, type EProcessStep, type EnsembleAggregate, type EnsembleJudgeOptions, type EquivalenceArm, type EquivalenceCheckDefinition, type EquivalenceCheckSpec, type EquivalenceChecker, type EquivalenceCheckerInput, type EquivalenceCheckerResult, type EquivalenceObligation, type EquivalenceObligationStatus, EquivalenceProtocolError, type EquivalenceRecord, type ErrorCluster, type ErrorCountPattern, type ErrorStreakOptions, type EvalCampaignOptions, type EvalCampaignResult, type EventFilter, type EvidenceRef, type ExactAnalystRunEvent, type ExactAnalystRunResult, type ExactCapableAnalyst, type ExactRegistryRunOpts, type ExactRiskDifferenceResult, type ExperimentPlan, type ExperimentRep, type ExperimentResult, type ExperimentStats, type ExtractOptions, type ExtractResult, type ExtractedUsage, FAILURE_CLASSES, FAILURE_MODE_KIND_SPEC, type FailedRun, type FailureBlame, type FailureClass, type FailureClassification, type FailureContext, type FailureMode, type FailureReasonRule, type FailureRule, type FeedbackArtifactType, type FeedbackAttempt, type FeedbackLabel, type FeedbackLabelKind, type FeedbackLabelSource, type FeedbackOptimizerRow, type FeedbackOutcome, type FeedbackSplitPolicy, type FeedbackTask, type FeedbackTrajectory, type FeedbackTrajectoryFilter, type FeedbackTrajectoryStore, type FieldDestination, FileSystemFeedbackTrajectoryStore, FileSystemRawProviderSink, FileSystemTraceStore, type Finding, type FindingSubject, type FindingSubjectKind, type FindingToPolicyEditOptions, type FindingsDiff, FindingsStore, type GainDistributionBin, type GainDistributionFigureSpec, type GainDistributionOptions, type GateDecision, type GateEvidence, type GenericSpan, type GoldenItem, HARNESS_NATIVE_MODEL, type HarnessType, type HeadroomInput, HeldOutGate, type HeldOutGateConfig, type HeldOutGateDecision, type HeldOutGateRejectionCode, type HeldOutPartition, type HiddenCriteriaGrader, type HiddenGradeResult, type HiddenLeak, type ImprovementThresholds, type ImprovementVerdictResult, InMemoryFeedbackTrajectoryStore, InMemoryRawProviderSink, InMemoryTraceStore, type InsightReport, type IntentMatchInput, type IntentMatchOptions, type IntentMatchResult, type InterimReleaseConfidence, type InterimReleaseConfidenceInput, type JudgeAgreement, type JudgeCalibrationSet, type JudgeConfusion, JudgeError, type JudgeExampleAgreement, type JudgeFamily, type JudgeFn, type JudgeGateDecision, type JudgeGateInput, type JudgeGatePolicy, JudgeGateRefusedError, type JudgeGateStatus, type JudgeInput, type JudgeRetryOutcome, type JudgeRetryPolicy, type JudgeRubric, type JudgeRun, type JudgeRunMeta, type JudgeScore, type JudgeScoreInput, type JudgeScoresRecord, type JudgeSpan, type JudgeVerdict, type KeywordConceptSpec, type KeywordCoverageFinding, type KeywordCoverageOptions, type KeywordCoverageResult, type KnowledgeAcquisitionMode, type KnowledgeBundle, type KnowledgeFreshness, type KnowledgeImportance, type KnowledgeReadinessReport, type KnowledgeRequirement, type KnowledgeRequirementCategory, type KnowledgeSensitivity, type Layer, type LayerResult, type LayerStatus, type LeaderboardOptions, type LeaderboardRow, type LinearRegressionResult, type LlmCallMetadata, type LlmCallRequest, type LlmCallResult, type LlmChargeBounds, type LlmCorrectnessCheckerOpts, type LlmJudgeOptions, type LlmMessage, type LlmReviewerConfig, type LlmSpan, type LlmTokenLogprob, type LlmToolCall, type LlmToolChoice, type LlmToolDefinition, type LlmUsage, type MannWhitneyResult, type MatchedPair, type MatchedRunRecordPair, type MaximumCharge, type McNemarResult, type MintRolloutOptions, type MintRolloutResult, type MintedRolloutLine, type MockTransportOpts, type ModelEndpointCheck, type ModelEndpointRequest, type ModelSeats, ModelSubstitutionError, MultiLayerVerifier, type NoLeakOptions, NoopRawProviderSink, NotFoundError, OUTPUT_VALUE, type Objective, type OpenAiCompatibleTransportOpts, type Oracle, type OracleObservation, type OracleReport, type OracleResult, type OrderMode, type OtlpExport, OtlpFileTraceStore, type OtlpFlatLine, type OtlpSpan, type OwnerVerdict, type PaidCallResult, type PairArmsOptions, type PairArmsResult, type PairRunRecordsResult, type PairedArmRow, type PairedArmsComparison, type PairedBootstrapOptions, type PairedBootstrapResult, type PairedCorrectness, type PairedDecisionMethod, type PairedDecisionShape, type PairedDecisionStatistic, type PairedDeltaTestOptions, type PairedDeltaTestResult, type PairedEvalueOptions, type PairedEvalueSequence, type PairedEvalueStep, type PairedMcNemarEvidence, type PairedMetricDelta, type PairedPromotionAlternative, type PairedPromotionAlternativeCell, type PairedPromotionCallRates, type PairedPromotionDecision, type PairedPromotionDecisionOptions, type PairedPromotionPowerCall, type PairedPromotionPowerOptions, type PairedPromotionPowerResult, type PairedSignTestResult, type PairedTTestResult, type PairedValueLaw, type ParetoPoint, type ParetoResult, type PartitionHeldOutOptions, type PendingCostCall, type PendingCostCallView, type PersistedFinding, type PersonaConfig, type PolicyEdit, type PolicyEditAdmission, type PolicyEditAdmissionOptions, type PolicyEditAxis, type PolicyEditCandidateRecord, type PolicyEditChange, type PolicyEditExpectedGain, type PolicyEditGainDirection, type PolicyEditGainUnit, type PolicyEditInit, type PolicyEditRisk, type PolicyEditSchemaVersion, type PolicyEditSource, type PolicyEditTarget, type PolicyEditTargetSurface, type PreferenceMemoryEntry, type ProducedProposal, type ProducedState, type ProductBenchmarkExportOptions, type ProductBenchmarkExportResult, type ProductBenchmarkManifest, type ProductBenchmarkSingleRunExportOptions, type ProductBenchmarkSplit, type ProductBenchmarkValidationReport, type ProductClientConfig, type ProductScenario, type ProfileAxisSpec, type ProjectRuntimeTrajectoryEvidenceOptions, type PromptHandle, type ProportionInterval, type ProposalEventLike, type ProposalFinding, type ProposeFn, type ProposeInput, type ProposeOutput, type ProposeReviewConfig, type ProposeReviewControlAction, type ProposeReviewControlConfig, type ProposeReviewControlResult, type ProposeReviewControlState, type ProposeReviewReport, type ProposedSideEffect, type QueryTracesPage, REDACTION_VERSION, type RankTestMethod, type RankTestMethodRequest, type RankTestOptions, type RawAnalystFinding, type RawProviderEvent, type RawProviderSink, type RecordRunsOptions, type RedTeamCase, type RedTeamCategory, type RedTeamFinding, type RedTeamReport, type RedactOptions, type RedactionProfile, type RedactionReport, type ReferenceReplayCaseRun, type ReferenceReplayRun, type ReferenceReplaySplit, type ReflectionContext, type ReflectionProposal, type RegistryRunOpts, type ReleaseConfidenceInput, type ReleaseConfidenceIssue, type ReleaseConfidenceMetrics, type ReleaseConfidenceScorecard, type ReleaseTraceEvidence, type RepeatedActionOptions, type RequiredPairsForPairedPromotionOptions, type RequiredPairsForPairedPromotionResult, type RequirementCheck, type ResearchReport, type ResearchReportOptions, type Researcher, type RetryWrite, type Review, type ReviewFn, type ReviewInput, type ReviewMemoryEntry, type ReviewMemoryStore, type RiskDifferenceResult, type RolloutSearchLineage, type RouteMap, type RoutedField, type RubricDimension, type Run, type RunCampaignOptions, type RunCaptureReport, type RunCommandInput, type RunCommandResult, type RunCostProvenance, type RunFilter, RunIntegrityError, type RunIntegrityReport, type RunJudgeMetadata, type RunOutcome, type RunPaidCallInput, type RunRecord, type RunRecordBackend, RunRecordValidationError, type RunScore, type RunScoreWeights, type RunSearchCoordinates, type RunSplitTag, type RunStatus, type RunTaskFailure, type RunTerminalOutcome, type RunTokenUsage, type RunTraceRef, type RuntimeEventLike, type RuntimeTrajectoryEvidenceProjection, type RuntimeTrajectoryEvidenceSummary, type RuntimeTrajectoryHookEvent, type RuntimeTrajectoryRecord, type RuntimeTrajectoryRunRecord, SEMANTIC_CONCEPT_JUDGE_VERSION, type SampleSummary, type SampleSummaryOptions, type SandboxDriver, type SandboxSdkTransportOpts, type SatisfiedBy, type Scenario, type ScenarioCost, type ScenarioFile, type ScenarioResult, type ScorePreference, type ScoreRiskDifferenceResult, type Scorecard, type ScorecardCell, type ScorecardCellDiff, type ScorecardDiff, type ScorecardEntry, type ScorecardLogLine, type SearchSpanResult, type SearchTraceResult, type SelfImproveBudget, type SelfImproveMethodOptions, type SelfImproveMethodProvenance, type SelfImproveMethodResult, type SelfImproveOptions, type SelfImproveProgressEvent, type SelfImproveProposerOptions, type SelfImproveProposerResult, type SelfImproveResult, type SemanticConceptJudgeInput, type SemanticConceptJudgeOptions, type SemanticConceptJudgeResult, type SequentialDecision, type SeriesConvergenceOptions, type SeriesConvergenceResult, type SeriesDistribution, type ServedModelCheck, type ServedModelPolicy, type ServedModelVerdict, type Severity, type ShareSafetyStatus, type ShareSafetyVerdict, type SignTestAlternative, type SingleBackendDivergence, type SingleBackendReport, type Span, type SpanFilter, type SpanHandle, type SpanMatchRecord, type SpanPredicate, type SpanRecordsToTraceAnalysisStoreOptions, type SplitCoverage, type SteeringBundle, type SteeringChange, type SteeringOptimizationResult, type SteeringOptimizationRow, type StopDecision, type StrategyChecker, type StreamingDetector, type SummaryTable, type SummaryTableOptions, type SummaryTableRow, type SynthesisTarget, type TaskGold, type TaskHeadroom, type ToolCallEventLike, type ToolSpan, type ToolStats, type ToolUseMetrics, type ToolUseOptions, type TraceAnalysisEngine, type TraceAnalysisEngineResult, type TraceAnalysisStore, type TraceAnalystByteBudgets, type TraceAnalystDefinition, type TraceAnalystFilters, type TraceAnalystSpan, type TraceAnalystSpanKind, type TraceAnalystSpanStatus, type TraceAnalystTraceSummary, type TraceContract, type TraceContractSpec, TraceEmitter, type TraceEmitterOptions, type TraceEvent, type TraceInsightReadiness, type TraceInsightSuite, type TraceStore, type TrainSelectionTestOptions, type TrainSelectionTestPartition, type Trajectory, type TrajectoryStep, type TreatmentGate, type TreatmentGateInput, type TreatmentGateOptions, type TrialTrace, type Turn, type TurnResult, UNKNOWN_MODEL, type UserQuestion, VERIFICATION_STRATEGIES, VERIFICATION_STRATEGY_SOURCES, ValidationError, type VerbosityBiasResult, type Verdict, type VerdictCacheStore, type VerdictCertification, type Verification, type VerificationReport, type VerificationStrategyProfile, type VerificationStrategySource, type VerifyFn, type VerifyOptions, type ViewSpansResult, type ViewTraceOversized, type ViewTraceResult, type ViteDeployRunnerInput, type WeightedCompositeInput, type WeightedCompositeResult, type WelchTestResult, type WelchTestStatus, type WilcoxonSignedRankResult, type WranglerDeployRunnerInput, addCalibrationVerdicts, agentProfileCellKey, agentProfileHash, agentProfileId, analystRunDigest, analystRunToFeedbackTrajectory, analystRunToReviewRequests, analyzeRuns, analyzeSeries, analyzeTraces, argHash, assertCrossFamily, assertCrossFamilyServed, assertJudgeMayGate, assertRealBackend, assertRunCaptured, assertServedModel, assertServedModels, assessShareSafety, assignFeedbackSplit, attest, benjaminiHochberg, bonferroni, bootstrapCi, budgetBreachView, buildAgentProfileCell, buildDefaultAnalystRegistry, buildTraceInsightContext, buildTraceInsightPrompt, buildTrajectory, calibrateJudge, calibrateJudgeContinuous, captureFetchToRawSink, certificationEvidenceDigest, checkServedModel, classifyFailure, classifyFailureReason, comparePairedArms, completionVerdict, computeFindingId, computeToolUseMetrics, confidenceInterval, continuousAgreement, controlRunToFeedbackTrajectory, corpusInterRaterAgreement, corpusInterRaterAgreementFromJudgeScores, createBoundedTraceAnalysisStore, createChatClient, createChatTraceEngine, createDspyRlmTraceEngine, createFeedbackTrajectory, createLlmCorrectnessChecker, createTraceAnalyst, decidePairedPromotion, defineAgentEval, defineEquivalenceCheck, describeTraceInsightScope, diffFindings, domainEvidencePattern, dominates, eProcess, equivalenceVerdict, evaluateInterimReleaseConfidence, evaluateReleaseConfidence, expandProfileAxes, exportRunAsOtlp, extractProducedState, extractUsage, extractUsageFromSse, failureBlame, failureClusterView, feedbackTrajectoriesToDatasetScenarios, feedbackTrajectoriesToOptimizerRows, feedbackTrajectoryToOptimizerRow, gainHistogram, gradeSemanticStatus, groupRunsByAgentProfileCell, harnessAxisOf, hashJson, holm, inferDomainKeywords, interRaterReliability, iqr, isJudgeSpan, isLlmSpan, isRunRecord, isToolSpan, judgeAgreementView, judgeFamily, judgeGateDecision, judgeSpans, llmJudge, makeFinding, makeProposalFinding, manifestContentDigest, mannWhitneyU, mcnemar, mcnemarPower, mcnemarRequiredN, measureJudgeAgreement, mintRolloutRows, modelHasSnapshot, mulberry32, otlpTextToTraceAnalysisStore, pairArms, pairRunRecords, pairedBootstrap, pairedEvalueSequence, pairedMde, pairedRiskDifference, pairedRiskDifferenceExact, pairedRiskDifferenceScore, pairedSignTest, pairedTTest, paretoFrontier, parseRunRecordSafe, partialCredit, passAtK, pearsonR, index_d_exports as profile, ranks, redTeamDataset, redTeamReport, redact, redactForShare, redactText, registerCalibrationSet, renderPreferenceMemoryMarkdown, requiredPairedSampleSize, requiredSampleSize, runCampaign, runCanaries, runEquivalenceCheck, runEvalCampaign, runSemanticConceptJudge, runTaskScore, runsForScenario, scoreRedTeamOutput, scoreTraceInsightReadiness, selfImprove, shareAllowed, snapshotFromJudgeAgreement, spanRecordsToTraceAnalysisStore, spearmanR, summarizeBackendIntegrity, summarizeNumberSeries, summarizePreferenceMemory, summaryTable, toAgentProfileJson, tokenizeDomainWords, toolSpansToTraceAnalysisStore, toolWasteView, transientDispatchFailure, validateRunRecord, verbosityBias, verifyAgentProfileCell, verifyAttestation, verifyCompletion, weightedComposite, weightedMean, welchsTTest, wilcoxonSignedRank, wilson, withAssignedFeedbackSplit };
+export { AGENT_PROFILE_KINDS, ATTESTATION_ALGORITHM, type ActionExecutionPolicy, type ActionPolicyDecision, type ActionableSideInfo, type ActiveLearningOptions, type AgentEvalAgent, AgentEvalError, type AgentEvalErrorCode, type AgentEvalEvaluateOptions, type AgentEvalImproveOptions, type AgentProfile, type AgentProfileCell, type AgentProfileCellInput, type AgentProfileDimensionValue, type AgentProfileJson, type AgentProfileSourceInput, type Analyst, type AnalystContext, type AnalystFeedbackTrajectoryOptions, type AnalystFinding, type AnalystFindingDigest, AnalystRegistry, type AnalystRegistryOptions, type AnalystReviewDecision, type AnalystReviewRequest, type AnalystRunDigest, type AnalystRunEvent, type AnalystRunInputs, type AnalystRunResult, type AnalystRunSummary, type AnalystSeverity, type AnalystUsageReceipt, type AnalyzeRunsOptions, type AnalyzeTracesResult, type ArgumentCheck, type ArmAssignment, type ArmPopulationReading, type ArmPopulationSummary, type Artifact, type ArtifactCheck, type ArtifactEventLike, type ArtifactResult, type AssertCapabilityHeadroomOptions, type AssertCrossFamilyOptions, type AssertCrossFamilyServedOptions, type AssertServedModelOptions, type AssertSingleBackendOptions, type AttestationProvenance, type AttestationVerification, type AttestedReport, BOOTSTRAP_GATE_MIN_N, type BackendDescriptor, BackendIntegrityError, type BackendIntegrityReport, type BenchmarkEvaluation, type BlendWeights, type BootstrapOptions, type BootstrapResult, type BoundedProcessInput, type BoundedProcessResult, type BudgetLedgerEntry, type BudgetPolicy, type BudgetSpec, CODING_HARNESSES, type CalibrationExample, type CalibrationMerge, type CalibrationResult, type CampaignCellFailureReceipt, type CampaignCellRetryPolicy, type CampaignFactoryParams, type CampaignIntegrityPolicy, type CampaignResult, type CampaignRunContext, type CampaignRunOutcome, type CampaignScenario, type CampaignVariant, type CanaryAlert, type CanaryEvaluation, type CanaryKind, type CanaryLeak, type CanaryOptions, type CanaryReport, type CandidateScore, type CapabilityHeadroomOptions, type CapabilityHeadroomResult, type CaptureErrorHandler, type CaptureWrite, type CellVerdict, type ChannelRollup, type ChatCallOpts, type ChatClient, type ChatRequest, type ChatResponse, type ChatTraceEngineOptions, type ChatTransport, type CheckResult, type CheckerIdentity, type CheckerOutcome, type CliffsMagnitude, type CollectedArtifacts, type CommandRunner, type ComparePairedArmsOptions, type CompletionCriterion, type CompletionRequirement, type CompletionVerdict, type ConceptFinding, type ConceptSpec, ConfigError, type ContinuousAgreement, type ContinuousAgreementOptions, type ContinuousCalibrationResult, type ContractCheckResult, type ContractLintFinding, type ContractRule, type ContractRuleExecution, type ContractRun, type ContractSpan, type ContractStatus, type ContractVerdict, type ControlActionOutcome, type ControlBudget, type ControlContext, type ControlDecision, type ControlEvalResult, type ControlRunResult, type ControlRuntimeConfig, type ControlRuntimeError, type ControlStep, type CorpusAgreementOptions, type CorpusAgreementPerDimension, type CorpusAgreementReport, type CorpusScoreRecord, type CorrectnessChecker, type CostChannel, type CostEntry, CostLedger, type CostLedgerFilter, type CostLedgerHandle, type CostLedgerOptions, type CostLedgerPersistence, type CostLedgerSummary, type CostProvenance, type CostReceipt, type CostReceiptInput, type CostResult, type CostSummary, type CostUsage, type CounterfactualContext, type CounterfactualMutation, type CounterfactualResult, type CounterfactualRunner, type CreateChatClientOpts, type CreateTraceAnalystOptions, CrossFamilyError, type CustomTokenPricing, type CustomTransportOpts, DEFAULT_JUDGE_GATE_POLICY, DEFAULT_RED_TEAM_CORPUS, DEFAULT_TRACE_ANALYST_BUDGETS, DEFAULT_TRACE_ANALYST_KINDS, type DataAcquisitionPlan, type DatasetManifest, type DatasetOverview, type DatasetScenario, type DatasetSplit, type DefaultAnalystRegistryOptions, type DefaultVerdict, type DefineAgentEvalOptions, type DefinedAgentEval, type DeployGateLayerInput, type DeployRunResult, type DeployRunner, type DetectorEvent, type DetectorSignal, type DiffPolicy, type DiffScorecardOptions, type DimensionJudgeScore, type DirEntry, type DiscoverPersonasOptions, type DiscoveredPersona, type DriverState, type DspyRlmTraceEngineOptions, type EProcess, type EProcessOptions, type EProcessState, type EProcessStep, type EnsembleAggregate, type EnsembleJudgeOptions, type EquivalenceArm, type EquivalenceCheckDefinition, type EquivalenceCheckSpec, type EquivalenceChecker, type EquivalenceCheckerInput, type EquivalenceCheckerResult, type EquivalenceObligation, type EquivalenceObligationStatus, EquivalenceProtocolError, type EquivalenceRecord, type ErrorCluster, type ErrorCountPattern, type ErrorStreakOptions, type EvalCampaignOptions, type EvalCampaignResult, type EventFilter, type EvidenceRef, type ExactAnalystRunEvent, type ExactAnalystRunResult, type ExactCapableAnalyst, type ExactRegistryRunOpts, type ExactRiskDifferenceResult, type ExperimentPlan, type ExperimentRep, type ExperimentResult, type ExperimentStats, type ExtractOptions, type ExtractResult, type ExtractedUsage, FAILURE_CLASSES, FAILURE_MODE_KIND_SPEC, type FailedRun, type FailureBlame, type FailureClass, type FailureClassification, type FailureContext, type FailureMode, type FailureReasonRule, type FailureRule, type FeedbackArtifactType, type FeedbackAttempt, type FeedbackLabel, type FeedbackLabelKind, type FeedbackLabelSource, type FeedbackOptimizerRow, type FeedbackOutcome, type FeedbackSplitPolicy, type FeedbackTask, type FeedbackTrajectory, type FeedbackTrajectoryFilter, type FeedbackTrajectoryStore, type FieldDestination, FileSystemFeedbackTrajectoryStore, FileSystemRawProviderSink, FileSystemTraceStore, type Finding, type FindingSubject, type FindingSubjectKind, type FindingToPolicyEditOptions, type FindingsDiff, FindingsStore, type GainDistributionBin, type GainDistributionFigureSpec, type GainDistributionOptions, type GateDecision, type GateEvidence, type GenericSpan, type GoldenItem, HARNESS_NATIVE_MODEL, type HarnessType, type HeadroomInput, HeldOutGate, type HeldOutGateConfig, type HeldOutGateDecision, type HeldOutGateRejectionCode, type HeldOutPartition, type HiddenCriteriaGrader, type HiddenGradeResult, type HiddenLeak, type ImprovementThresholds, type ImprovementVerdictResult, InMemoryFeedbackTrajectoryStore, InMemoryRawProviderSink, InMemoryTraceStore, type InsightReport, type IntentMatchInput, type IntentMatchOptions, type IntentMatchResult, type InterimReleaseConfidence, type InterimReleaseConfidenceInput, type JudgeAgreement, type JudgeCalibrationSet, type JudgeConfusion, JudgeError, type JudgeExampleAgreement, type JudgeFamily, type JudgeFn, type JudgeGateDecision, type JudgeGateInput, type JudgeGatePolicy, JudgeGateRefusedError, type JudgeGateStatus, type JudgeInput, type JudgeRetryOutcome, type JudgeRetryPolicy, type JudgeRubric, type JudgeRun, type JudgeRunMeta, type JudgeScore, type JudgeScoreInput, type JudgeScoresRecord, type JudgeSpan, type JudgeVerdict, type KeywordConceptSpec, type KeywordCoverageFinding, type KeywordCoverageOptions, type KeywordCoverageResult, type KnowledgeAcquisitionMode, type KnowledgeBundle, type KnowledgeFreshness, type KnowledgeImportance, type KnowledgeReadinessReport, type KnowledgeRequirement, type KnowledgeRequirementCategory, type KnowledgeSensitivity, type Layer, type LayerResult, type LayerStatus, type LeaderboardOptions, type LeaderboardRow, type LinearRegressionResult, type LlmCallMetadata, type LlmCallRequest, type LlmCallResult, type LlmChargeBounds, type LlmCorrectnessCheckerOpts, type LlmJudgeOptions, type LlmMessage, type LlmReviewerConfig, type LlmSpan, type LlmTokenLogprob, type LlmToolCall, type LlmToolChoice, type LlmToolDefinition, type LlmUsage, type MannWhitneyResult, type MatchedPair, type MatchedRunRecordPair, type MaximumCharge, type McNemarResult, type MintRolloutOptions, type MintRolloutResult, type MintedRolloutLine, type MockTransportOpts, type ModelEndpointCheck, type ModelEndpointRequest, type ModelSeats, ModelSubstitutionError, MultiLayerVerifier, type NoLeakOptions, NoopRawProviderSink, NotFoundError, OUTPUT_VALUE, type Objective, type OpenAiCompatibleTransportOpts, type Oracle, type OracleObservation, type OracleReport, type OracleResult, type OrderMode, type OtlpExport, OtlpFileTraceStore, type OtlpFlatLine, type OtlpSpan, type OwnerVerdict, type PaidCallResult, type PairArmsOptions, type PairArmsResult, type PairRunRecordsResult, type PairedArmRow, type PairedArmsComparison, type PairedBootstrapOptions, type PairedBootstrapResult, type PairedCorrectness, type PairedDecisionMethod, type PairedDecisionShape, type PairedDecisionStatistic, type PairedDeltaTestOptions, type PairedDeltaTestResult, type PairedEvalueOptions, type PairedEvalueSequence, type PairedEvalueStep, type PairedMcNemarEvidence, type PairedMetricDelta, type PairedPromotionAlternative, type PairedPromotionAlternativeCell, type PairedPromotionCallRates, type PairedPromotionDecision, type PairedPromotionDecisionOptions, type PairedPromotionPowerCall, type PairedPromotionPowerOptions, type PairedPromotionPowerResult, type PairedSignTestResult, type PairedTTestResult, type PairedValueLaw, type ParetoPoint, type ParetoResult, type PartitionHeldOutOptions, type PendingCostCall, type PendingCostCallView, type PersistedFinding, type PersonaConfig, type PolicyEdit, type PolicyEditAdmission, type PolicyEditAdmissionOptions, type PolicyEditAxis, type PolicyEditCandidateRecord, type PolicyEditChange, type PolicyEditExpectedGain, type PolicyEditGainDirection, type PolicyEditGainUnit, type PolicyEditInit, type PolicyEditRisk, type PolicyEditSchemaVersion, type PolicyEditSource, type PolicyEditTarget, type PolicyEditTargetSurface, type PreferenceMemoryEntry, type ProducedProposal, type ProducedState, type ProductBenchmarkExportOptions, type ProductBenchmarkExportResult, type ProductBenchmarkManifest, type ProductBenchmarkSingleRunExportOptions, type ProductBenchmarkSplit, type ProductBenchmarkValidationReport, type ProductClientConfig, type ProductScenario, type ProfileAxisSpec, type ProjectRuntimeTrajectoryEvidenceOptions, type PromptHandle, type ProportionInterval, type ProposalEventLike, type ProposalFinding, type ProposeFn, type ProposeInput, type ProposeOutput, type ProposeReviewConfig, type ProposeReviewControlAction, type ProposeReviewControlConfig, type ProposeReviewControlResult, type ProposeReviewControlState, type ProposeReviewReport, type ProposedSideEffect, type QueryTracesPage, REDACTION_VERSION, type RankTestMethod, type RankTestMethodRequest, type RankTestOptions, type RawAnalystFinding, type RawProviderEvent, type RawProviderSink, type RecordRunsOptions, type RedTeamCase, type RedTeamCategory, type RedTeamFinding, type RedTeamReport, type RedactOptions, type RedactionProfile, type RedactionReport, type ReferenceReplayCaseRun, type ReferenceReplayRun, type ReferenceReplaySplit, type ReflectionContext, type ReflectionProposal, type RegistryRunOpts, type ReleaseConfidenceInput, type ReleaseConfidenceIssue, type ReleaseConfidenceMetrics, type ReleaseConfidenceScorecard, type ReleaseTraceEvidence, type RepeatedActionOptions, type RequiredPairsForPairedPromotionOptions, type RequiredPairsForPairedPromotionResult, type RequirementCheck, type ResearchReport, type ResearchReportOptions, type Researcher, type RetryWrite, type Review, type ReviewFn, type ReviewInput, type ReviewMemoryEntry, type ReviewMemoryStore, type RiskDifferenceResult, type RolloutSearchLineage, type RouteMap, type RoutedField, type RubricDimension, type Run, type RunCampaignOptions, type RunCaptureReport, type RunCommandInput, type RunCommandResult, type RunCostProvenance, type RunFilter, RunIntegrityError, type RunIntegrityReport, type RunJudgeMetadata, type RunOutcome, type RunPaidCallInput, type RunRecord, type RunRecordBackend, RunRecordValidationError, type RunScore, type RunScoreWeights, type RunSearchCoordinates, type RunSplitTag, type RunStatus, type RunTaskFailure, type RunTerminalOutcome, type RunTokenUsage, type RunTraceRef, type RuntimeEventLike, type RuntimeTrajectoryEvidenceProjection, type RuntimeTrajectoryEvidenceSummary, type RuntimeTrajectoryHookEvent, type RuntimeTrajectoryRecord, type RuntimeTrajectoryRunRecord, SEMANTIC_CONCEPT_JUDGE_VERSION, type SampleSummary, type SampleSummaryOptions, type SandboxDriver, type SandboxSdkTransportOpts, type SatisfiedBy, type Scenario, type ScenarioCost, type ScenarioFile, type ScenarioResult, type ScorePreference, type ScoreRiskDifferenceResult, type Scorecard, type ScorecardCell, type ScorecardCellDiff, type ScorecardDiff, type ScorecardEntry, type ScorecardLogLine, type SearchSpanResult, type SearchTraceResult, type SelfImproveBudget, type SelfImproveMethodOptions, type SelfImproveMethodProvenance, type SelfImproveMethodResult, type SelfImproveOptions, type SelfImproveProgressEvent, type SelfImproveProposerOptions, type SelfImproveProposerResult, type SelfImproveResult, type SemanticConceptJudgeInput, type SemanticConceptJudgeOptions, type SemanticConceptJudgeResult, type SequentialDecision, type SeriesConvergenceOptions, type SeriesConvergenceResult, type SeriesDistribution, type ServedModelCheck, type ServedModelPolicy, type ServedModelVerdict, type Severity, type ShareSafetyStatus, type ShareSafetyVerdict, type SignTestAlternative, type SingleBackendDivergence, type SingleBackendReport, type Span, type SpanFilter, type SpanHandle, type SpanMatchRecord, type SpanPredicate, type SpanRecordsToTraceAnalysisStoreOptions, type SplitCoverage, type SteeringBundle, type SteeringChange, type SteeringOptimizationResult, type SteeringOptimizationRow, type StopDecision, type StrategyChecker, type StreamingDetector, type SummaryTable, type SummaryTableOptions, type SummaryTableRow, type SynthesisTarget, type TaskGold, type TaskHeadroom, type ToolCallEventLike, type ToolSpan, type ToolStats, type ToolUseMetrics, type ToolUseOptions, type TraceAnalysisEngine, type TraceAnalysisEngineResult, type TraceAnalysisStore, type TraceAnalystByteBudgets, type TraceAnalystDefinition, type TraceAnalystFilters, type TraceAnalystSpan, type TraceAnalystSpanKind, type TraceAnalystSpanStatus, type TraceAnalystTraceSummary, type TraceContract, type TraceContractSpec, TraceEmitter, type TraceEmitterOptions, type TraceEvent, type TraceInsightReadiness, type TraceInsightSuite, type TraceStore, type TrainSelectionTestOptions, type TrainSelectionTestPartition, type Trajectory, type TrajectoryStep, type TreatmentGate, type TreatmentGateInput, type TreatmentGateOptions, type TrialTrace, type Turn, type TurnResult, UNKNOWN_MODEL, type UserQuestion, VERIFICATION_STRATEGIES, VERIFICATION_STRATEGY_SOURCES, ValidationError, type VerbosityBiasResult, type Verdict, type VerdictCacheStore, type VerdictCertification, type Verification, type VerificationReport, type VerificationStrategyProfile, type VerificationStrategySource, type VerifyFn, type VerifyOptions, type ViewSpansResult, type ViewTraceOversized, type ViewTraceResult, type ViteDeployRunnerInput, type WeightedCompositeInput, type WeightedCompositeResult, type WelchTestResult, type WelchTestStatus, type WilcoxonSignedRankResult, type WranglerDeployRunnerInput, addCalibrationVerdicts, agentProfileCellKey, agentProfileHash, agentProfileId, analystRunDigest, analystRunToFeedbackTrajectory, analystRunToReviewRequests, analyzeRuns, analyzeSeries, analyzeTraces, argHash, assertCrossFamily, assertCrossFamilyServed, assertJudgeMayGate, assertRealBackend, assertRunCaptured, assertServedModel, assertServedModels, assessShareSafety, assignFeedbackSplit, attest, benjaminiHochberg, bonferroni, bootstrapCi, budgetBreachView, buildAgentProfileCell, buildDefaultAnalystRegistry, buildTraceInsightContext, buildTraceInsightPrompt, buildTrajectory, calibrateJudge, calibrateJudgeContinuous, captureFetchToRawSink, certificationEvidenceDigest, checkServedModel, classifyFailure, classifyFailureReason, comparePairedArms, completionVerdict, computeFindingId, computeToolUseMetrics, confidenceInterval, continuousAgreement, controlRunToFeedbackTrajectory, corpusInterRaterAgreement, corpusInterRaterAgreementFromJudgeScores, createBoundedTraceAnalysisStore, createChatClient, createChatTraceEngine, createDspyRlmTraceEngine, createFeedbackTrajectory, createLlmCorrectnessChecker, createTraceAnalyst, decidePairedPromotion, defineAgentEval, defineEquivalenceCheck, describeServedModelCheck, describeTraceInsightScope, diffFindings, domainEvidencePattern, dominates, eProcess, equivalenceVerdict, evaluateInterimReleaseConfidence, evaluateReleaseConfidence, expandProfileAxes, exportRunAsOtlp, extractProducedState, extractUsage, extractUsageFromSse, failureBlame, failureClusterView, feedbackTrajectoriesToDatasetScenarios, feedbackTrajectoriesToOptimizerRows, feedbackTrajectoryToOptimizerRow, gainHistogram, gradeSemanticStatus, groupRunsByAgentProfileCell, harnessAxisOf, hashJson, holm, inferDomainKeywords, interRaterReliability, iqr, isJudgeSpan, isLlmSpan, isRunRecord, isToolSpan, judgeAgreementView, judgeFamily, judgeGateDecision, judgeSpans, llmJudge, makeFinding, makeProposalFinding, manifestContentDigest, mannWhitneyU, mcnemar, mcnemarPower, mcnemarRequiredN, measureJudgeAgreement, mintRolloutRows, modelHasSnapshot, mulberry32, otlpTextToTraceAnalysisStore, pairArms, pairRunRecords, pairedBootstrap, pairedEvalueSequence, pairedMde, pairedRiskDifference, pairedRiskDifferenceExact, pairedRiskDifferenceScore, pairedSignTest, pairedTTest, paretoFrontier, parseRunRecordSafe, partialCredit, passAtK, pearsonR, index_d_exports as profile, ranks, redTeamDataset, redTeamReport, redact, redactForShare, redactText, registerCalibrationSet, renderPreferenceMemoryMarkdown, requiredPairedSampleSize, requiredSampleSize, runCampaign, runCanaries, runEquivalenceCheck, runEvalCampaign, runSemanticConceptJudge, runTaskScore, runsForScenario, scoreRedTeamOutput, scoreTraceInsightReadiness, selfImprove, shareAllowed, snapshotFromJudgeAgreement, spanRecordsToTraceAnalysisStore, spearmanR, summarizeBackendIntegrity, summarizeNumberSeries, summarizePreferenceMemory, summaryTable, toAgentProfileJson, tokenizeDomainWords, toolSpansToTraceAnalysisStore, toolWasteView, transientDispatchFailure, validateRunRecord, verbosityBias, verifyAgentProfileCell, verifyAttestation, verifyCompletion, weightedComposite, weightedMean, welchsTTest, wilcoxonSignedRank, wilson, withAssignedFeedbackSplit };
 //# sourceMappingURL=index.d.ts.map

package/dist/index.js

@@ -1,7 +1,7 @@
 import { t as __exportAll } from "./rolldown-runtime-8H4AJuhK.js";
 import { i as JudgeError, n as CaptureIntegrityError, o as NotFoundError, r as ConfigError, s as ValidationError, t as AgentEvalError } from "./errors-BcOs5MKW.js";
 import { a as hashCanonical, r as canonicalString } from "./canonical-Dq2hOQ8L.js";
-import { a as HARNESS_NATIVE_MODEL, c as agentProfileModelId, i as CODING_HARNESSES, l as expandProfileAxes, n as createLlmCorrectnessChecker, o as agentProfileHash, r as verifyCompletion, s as agentProfileId, t as completionVerdict, u as harnessAxisOf } from "./completion-verifier-YwoeujPl.js";
+import { a as HARNESS_NATIVE_MODEL, c as agentProfileModelId, i as CODING_HARNESSES, l as expandProfileAxes, n as createLlmCorrectnessChecker, o as agentProfileHash, r as verifyCompletion, s as agentProfileId, t as completionVerdict, u as harnessAxisOf } from "./completion-verifier-asaV1nSj.js";
 import { a as groupRunsByAgentProfileCell, c as verifyAgentProfileCell, i as buildAgentProfileCell, l as hashJson, o as toAgentProfileJson, r as agentProfileCellKey, s as validateAgentProfileCell, t as AGENT_PROFILE_KINDS, u as manifestContentDigest } from "./agent-profile-cell-B1V1gUZs.js";
 import { r as observedSplitScore, t as isRealnessGated } from "./reward-nw2xZGZG.js";
 import { i as isToolSpan, n as isJudgeSpan, r as isLlmSpan, t as FAILURE_CLASSES } from "./schema-Bt2vLFm3.js";
@@ -23,7 +23,7 @@ import { t as eProcess } from "./sequential-eprocess-DodI6RoF.js";
 import { i as pairedDeltaTest, n as pairedDecisionShape, r as minimumPairsForPairedDeltaTest, t as decidePairedPromotion } from "./paired-promotion-decision-D0SHb-ZL.js";
 import { c as runsForScenario, i as judgeSpans, t as argHash } from "./query-B_RZ36an.js";
 import { n as contentHash, r as fileVerdictCache, t as canonicalJson } from "./verdict-cache-CL_7tePD.js";
-import { G as runCanaries, H as redTeamDataset, Q as runCampaign, U as redTeamReport, V as DEFAULT_RED_TEAM_CORPUS, W as scoreRedTeamOutput, a as aggregateRunScore, i as transientDispatchFailure, o as clamp01, t as llmJudge } from "./llm-judge-3uAeRawE.js";
+import { G as runCanaries, H as redTeamDataset, Q as runCampaign, U as redTeamReport, V as DEFAULT_RED_TEAM_CORPUS, W as scoreRedTeamOutput, a as aggregateRunScore, i as transientDispatchFailure, o as clamp01, t as llmJudge } from "./llm-judge-Cybq29bL.js";
 import { a as resolveModelPricing, i as isModelPriced, n as estimateCost, r as estimateTokens, t as MODEL_PRICING } from "./metrics-BKaPBtIG.js";
 import { c as costForTokenPricing, i as CostCeilingReachedError, l as costForUsage, n as CostAccountingIncompleteError, t as CostLedger } from "./cost-ledger-BGZB6yY2.js";
 import { D as surfaceContentHash, M as assertRealBackend, N as summarizeBackendIntegrity, _ as parseReflectionResponse, a as ATTESTATION_ALGORITHM, g as buildReflectionPrompt, j as BackendIntegrityError, o as attest, s as verifyAttestation } from "./campaign-evidence-DunzcJGv.js";
@@ -31,10 +31,10 @@ import { a as redact, c as shareAllowed, n as assessShareSafety, o as redactForS
 import { a as makeProposalFinding, i as makeFinding, r as computeFindingId } from "./types-CnWmj8NW.js";
 import { S as requiredPairsForPairedPromotion, n as paretoFrontier, t as dominates, x as pairedPromotionPower } from "./pareto-BiBnQXAE.js";
 import { a as LlmCallError, i as stripFencedJson, n as isRecord, o as LlmResponseError } from "./json-BhdsJaOQ.js";
-import { r as selfImprove, t as defineAgentEval } from "./define-agent-eval-CJhECJNL.js";
+import { r as selfImprove, t as defineAgentEval } from "./define-agent-eval-CCbT7Kfr.js";
 import { n as equivalenceVerdict, t as certificationEvidenceDigest } from "./verdict-BGQmxQ8A.js";
 import { c as createBoundedTraceAnalysisStore, p as DEFAULT_TRACE_ANALYST_BUDGETS, t as createTraceAnalyst } from "./kind-factory-Dl0PmNQL.js";
-import { C as snapshotAnalystRun, S as readAnalystReview, b as assertUniqueFindingIds, c as FAILURE_MODE_KIND_SPEC, i as DEFAULT_TRACE_ANALYST_KINDS, n as AnalystRegistry, t as buildDefaultAnalystRegistry, v as analystFindingDigest, w as validateAnalystReviewDecisions, x as completedAnalystReviewQuality, y as analystRunDigest } from "./default-registry-BZ5FyYl9.js";
+import { C as snapshotAnalystRun, S as readAnalystReview, b as assertUniqueFindingIds, c as FAILURE_MODE_KIND_SPEC, i as DEFAULT_TRACE_ANALYST_KINDS, n as AnalystRegistry, t as buildDefaultAnalystRegistry, v as analystFindingDigest, w as validateAnalystReviewDecisions, x as completedAnalystReviewQuality, y as analystRunDigest } from "./default-registry-BBTDgkDM.js";
 import { OUTPUT_VALUE } from "./trace-attributes.js";
 import { n as extractUsageFromSse, t as extractUsage } from "./extract-usage-BZLU01kR.js";
 import { n as InMemoryRawProviderSink, r as NoopRawProviderSink, t as FileSystemRawProviderSink } from "./raw-provider-sink-75w4FWBj.js";
@@ -50,15 +50,15 @@ import { t as runEvalCampaign } from "./eval-campaign-DoX5vz3E.js";
 import "./rollout-Dd1ucT_M.js";
 import { t as mintRolloutRows } from "./mint-CbdwjTPW.js";
 import { n as pairedEvalueSequence, t as evaluateInterimReleaseConfidence } from "./sequential-BVv1sjsw.js";
-import { a as maximumChargeForLlmRequest, c as ModelSubstitutionError, f as assertServedModels, g as judgeFamily, h as assertCrossFamily, i as costReceiptFromLlmError, l as assertCrossFamilyServed, m as CrossFamilyError, o as backoffMs, p as checkServedModel, r as costReceiptFromLlm, s as isTransientLlmError, u as assertServedModel } from "./llm-client-DXuOTseU.js";
-import { n as paidJsonChat } from "./chat-json-call-BrZKn0L3.js";
-import { t as createDspyRlmTraceEngine } from "./dspy-rlm-engine-BTy7u0jQ.js";
-import { n as diffFindings, r as createChatTraceEngine, t as FindingsStore } from "./findings-store-DRiFdoid.js";
+import { _ as judgeFamily, a as maximumChargeForLlmRequest, c as ModelSubstitutionError, f as assertServedModels, g as assertCrossFamily, h as CrossFamilyError, i as costReceiptFromLlmError, l as assertCrossFamilyServed, m as describeServedModelCheck, p as checkServedModel, r as costReceiptFromLlm, s as isTransientLlmError, u as assertServedModel } from "./llm-client-b1ysZVIU.js";
+import { n as paidJsonChat } from "./chat-json-call-BzWYp18J.js";
+import { t as createDspyRlmTraceEngine } from "./dspy-rlm-engine-DN1p_5jx.js";
+import { n as diffFindings, r as createChatTraceEngine, t as FindingsStore } from "./findings-store-Bg4sdS-e.js";
 import { a as budgetBreachView, i as failureClusterView, n as computeToolUseMetrics, r as judgeAgreementView, t as toolWasteView } from "./tool-waste-DLsjS_p-.js";
-import { a as judgeGateDecision, c as snapshotFromJudgeAgreement, g as analyzeSeries, i as assertJudgeMayGate, n as JudgeGateRefusedError, o as measureJudgeAgreement, r as addCalibrationVerdicts, s as registerCalibrationSet, t as DEFAULT_JUDGE_GATE_POLICY } from "./judge-gate-Co5tuh57.js";
+import { _ as analyzeSeries, a as judgeGateDecision, c as snapshotFromJudgeAgreement, g as withJudgeRetry, i as assertJudgeMayGate, n as JudgeGateRefusedError, o as measureJudgeAgreement, r as addCalibrationVerdicts, s as registerCalibrationSet, t as DEFAULT_JUDGE_GATE_POLICY } from "./judge-gate-BVvmP00P.js";
 import { a as spanRecordsToTraceAnalysisStore, i as otlpTextToTraceAnalysisStore, r as OtlpFileTraceStore } from "./store-otlp-CFf27E8_.js";
 import { n as bootstrapCi, t as evaluateReleaseConfidence } from "./release-confidence-CYIFH-fp.js";
-import { t as createChatClient } from "./chat-client-CMQ8_h4J.js";
+import { t as createChatClient } from "./chat-client-BNATB3rJ.js";
 import { createHash } from "node:crypto";
 import { accessSync, appendFileSync, constants, cpSync, existsSync, mkdirSync, promises, readFileSync, readdirSync, statSync, writeFileSync } from "node:fs";
 import { basename, delimiter, dirname, extname, isAbsolute, join, relative, resolve } from "node:path";
@@ -7464,85 +7464,6 @@ function aggregateJudgeVerdicts(verdicts, dimensionKeys, weights) {
 	};
 }
 //#endregion
-//#region src/judge-retry.ts
-/**
-* Wrap a single judge LLM call with retry, optional fallback-model
-* rotation, exponential backoff, and a typed `JudgeRetryOutcome`. Callers
-* MUST inspect `succeeded` before using `value`; on failure the library
-* returns `value: null` rather than substituting a default, so a judge
-* abort cannot silently corrupt a downstream composite.
-*
-* Reporting contract: callers ship `TrialResult.judgeSucceeded = succeeded`
-* and `TrialResult.judgeAttempts = attempts` so `aggregateTrialsByMode`
-* with `mode: 'exclude-failed'` drops the trial.
-*/
-const DEFAULT_MAX_ATTEMPTS = 3;
-const DEFAULT_TIMEOUT_MS = 3e5;
-function sleep(ms) {
-	return new Promise((resolve) => setTimeout(resolve, ms));
-}
-/**
-* Wrap a judge call with retry + fallback-model + typed outcome semantics.
-*
-* The `judgeFn` signature is `(model: string, signal: AbortSignal) => Promise<T>`.
-* The signal will be aborted at `timeoutMs`. Callers should pass the signal
-* to their underlying fetch/SDK call so the abort actually fires.
-*
-* Returns a typed outcome — callers MUST inspect `succeeded` before using
-* `value`. The library refuses to default to a silent zero score because a
-* synthetic zero is indistinguishable from a real low score downstream.
-*/
-async function withJudgeRetry(judgeFn, policy = {}) {
-	const maxAttempts = policy.maxAttempts ?? DEFAULT_MAX_ATTEMPTS;
-	const timeoutMs = policy.timeoutMs ?? DEFAULT_TIMEOUT_MS;
-	const backoff = policy.backoffMs ?? backoffMs;
-	const isRetryable = policy.isRetryable ?? isTransientLlmError;
-	const models = policy.models && policy.models.length > 0 ? policy.models : [void 0];
-	let totalAttempts = 0;
-	const attemptErrors = [];
-	let lastError;
-	for (const model of models) for (let attempt = 0; attempt < maxAttempts; attempt++) {
-		totalAttempts += 1;
-		const controller = new AbortController();
-		const timer = setTimeout(() => controller.abort(/* @__PURE__ */ new Error("TimeoutError")), timeoutMs);
-		try {
-			const value = await judgeFn(model, controller.signal);
-			clearTimeout(timer);
-			return {
-				value,
-				succeeded: true,
-				attempts: totalAttempts,
-				modelUsed: model,
-				attemptErrors
-			};
-		} catch (err) {
-			clearTimeout(timer);
-			const errObj = err instanceof Error ? err : new Error(String(err));
-			lastError = errObj;
-			attemptErrors.push({
-				attempt: totalAttempts,
-				model: model ?? "(default)",
-				error: errObj.message
-			});
-			if (!isRetryable(errObj)) return {
-				value: null,
-				succeeded: false,
-				attempts: totalAttempts,
-				error: errObj,
-				attemptErrors
-			};
-			if (attempt < maxAttempts - 1) await sleep(backoff(attempt));
-		}
-	}
-	return {
-		value: null,
-		succeeded: false,
-		attempts: totalAttempts,
-		error: lastError,
-		attemptErrors
-	};
-}
-//#endregion
 //#region src/judge-panel.ts
 /**
 * Build a campaign-shaped `JudgeConfig` whose `score()` runs every panel
@@ -8955,6 +8876,6 @@ function rankRows(rows, weights) {
 	})).sort((a, b) => b.mean - a.mean);
 }
 //#endregion
-export { AGENT_PROFILE_KINDS, ATTESTATION_ALGORITHM, AgentEvalError, AnalystRegistry, BOOTSTRAP_GATE_MIN_N, BackendIntegrityError, BudgetBreachError, BudgetGuard, CODING_HARNESSES, ConfigError, CostAccountingIncompleteError, CostCeilingReachedError, CostLedger, CostTracker, CrossFamilyError, DEFAULT_BOUNDED_PROCESS_TIMEOUT_MS, DEFAULT_JUDGE_GATE_POLICY, DEFAULT_RED_TEAM_CORPUS, DEFAULT_TRACE_ANALYST_BUDGETS, DEFAULT_TRACE_ANALYST_KINDS, ERROR_COUNT_PATTERNS, EquivalenceProtocolError, FAILURE_CLASSES, FAILURE_MODE_KIND_SPEC, FileSystemFeedbackTrajectoryStore, FileSystemRawProviderSink, FileSystemTraceStore, FindingsStore, HARNESS_NATIVE_MODEL, HeldOutGate, InMemoryFeedbackTrajectoryStore, InMemoryRawProviderSink, InMemoryTraceStore, JudgeError, JudgeGateRefusedError, LlmCallError, LlmResponseError, MODEL_PRICING, ModelSubstitutionError, MultiLayerVerifier, NoopRawProviderSink, NotFoundError, OUTPUT_VALUE, OtlpFileTraceStore, PairwiseSteeringOptimizer, ProductClient, PromptRegistry, REDACTION_VERSION, RunIntegrityError, RunRecordValidationError, SEMANTIC_CONCEPT_JUDGE_VERSION, TraceEmitter, UNKNOWN_MODEL, VERIFICATION_STRATEGIES, VERIFICATION_STRATEGY_SOURCES, ValidationError, acquisitionPlansForKnowledgeGaps, addCalibrationVerdicts, agentProfileCellKey, agentProfileHash, agentProfileId, aggregateJudgeVerdicts, aggregateRunScore, analystRunDigest, analystRunToFeedbackTrajectory, analystRunToReviewRequests, analyzeAntiSlop, analyzeRuns, analyzeSeries, analyzeTraces, argHash, assertCapabilityHeadroom, assertCrossFamily, assertCrossFamilyServed, assertJudgeMayGate, assertNoHiddenLeak, assertProductBenchmarkRun, assertRealBackend, assertRunCaptured, assertServedModel, assertServedModels, assertSingleBackend, assessShareSafety, assignFeedbackSplit, attest, benjaminiHochberg, blendHeldout, blockingKnowledgeEval, bonferroni, bootstrapCi, budgetBreachView, buildAgentProfileCell, buildDefaultAnalystRegistry, buildReflectionPrompt, buildTraceInsightContext, buildTraceInsightPrompt, buildTrajectory, calibrateJudge, calibrateJudgeContinuous, canonicalJson, capabilityHeadroom, captureFetchToRawSink, certificationEvidenceDigest, checkCanaries, checkServedModel, checkTraceContracts, chiSquared, chiSquaredPValue, clamp01, classifyFailure, classifyFailureReason, cliffsDelta, cohensD, comparePairedArms, compileTraceContractSpec, completionVerdict, computeFindingId, computeToolUseMetrics, confidenceInterval, contentHash, continuousAgreement, controlRunToFeedbackTrajectory, corpusInterRaterAgreement, corpusInterRaterAgreementFromJudgeScores, costForTokenPricing, costForUsage, costReceiptFromLlm, costReceiptFromLlmError, createAntiSlopJudge, createBoundedTraceAnalysisStore, createChatClient, createChatTraceEngine, createDspyRlmTraceEngine, createFeedbackTrajectory, createLlmCorrectnessChecker, createLlmReviewer, createTraceAnalyst, decidePairedPromotion, defineAgentEval, defineEquivalenceCheck, deployGateLayer, describeTraceInsightScope, diffFindings, diffScorecard, discoverPersonas, domainEvidencePattern, dominates, eProcess, ensembleJudge, equivalenceVerdict, errorStreakDetector, estimateCost, estimateTokens, evaluateActionPolicy, evaluateInterimReleaseConfidence, evaluateOracles, evaluateReleaseConfidence, evaluateTraceContract, expandProfileAxes, explainTraceContract, exportProductBenchmark, exportProductBenchmarkRuns, exportRunAsOtlp, extractErrorCount, extractProducedState, extractUsage, extractUsageFromSse, failureBlame, failureClusterView, feedbackTrajectoriesToDatasetScenarios, feedbackTrajectoriesToOptimizerRows, feedbackTrajectoryToOptimizerRow, fileVerdictCache, fisherExactTwoSided, formatScorecardDiff, gainHistogram, gateTreatmentApplied, gradeSemanticStatus, groupRunsByAgentProfileCell, harnessAxisOf, hashContent, hashJson, holm, inMemoryReviewStore, inferDomainKeywords, interRaterReliability, interpretCliffs, iqr, isJudgeSpan, isLlmSpan, isModelPriced, isRunRecord, isToolSpan, isTransientLlmError, jsonShape, jsonlReviewStore, jsonlRunRecordBackend, judgeAgreementView, judgeFamily, judgeGateDecision, judgeSpans, knowledgeReadinessTracePayload, leaderboard, linearRegressionSlope, lintTraceContractSpec, llmJudge, loadScorecard, localCommandRunner, makeFinding, makePolicyEditCandidateRecord, makeProposalFinding, manifestContentDigest, mannWhitneyU, maximumChargeForLlmRequest, mcnemar, mcnemarPower, mcnemarRequiredN, measureJudgeAgreement, minimumPairsForPairedDeltaTest, mintRolloutRows, modelHasSnapshot, mulberry32, normalTwoSidedPValue, objectiveEval, observeAll, otlpTextToTraceAnalysisStore, pairArms, pairRunRecords, pairedBootstrap, pairedCohensDz, pairedDeltaTest, pairedEvalueSequence, pairedMde, pairedPromotionPower, pairedRiskDifference, pairedRiskDifferenceExact, pairedRiskDifferenceScore, pairedSignTest, pairedTTest, paretoFrontier, parseReflectionResponse, parseRunRecordSafe, partialCredit, partitionHeldOut, passAtK, pearsonR, preflightModels, productBenchmarkRepoIdentity, profile_exports as profile, projectRuntimeTrajectoryEvidence, proposeSynthesisTargets, ranks, readProductBenchmarkManifest, recordRuns, recordRunsToScorecard, redTeamDataset, redTeamReport, redact, redactForShare, redactText, registerCalibrationSet, renderPreferenceMemoryMarkdown, repeatedActionDetector, requiredPairedSampleSize, requiredPairsForPairedPromotion, requiredSampleSize, resolveModelPricing, resolveSeat, runAgentControlLoop, runBoundedProcess, runCampaign, runCanaries, runEquivalenceCheck, runEvalCampaign, runIntentMatchJudge, runKeywordCoverageJudge, runKeywordCoverageJudgeUrl, runProposeReview, runSemanticConceptJudge, runTaskScore, runsForScenario, sampleMedian, scoreKnowledgeReadiness, scoreRedTeamOutput, scoreTraceInsightReadiness, seatPresets, selfImprove, shareAllowed, snapshotFromJudgeAgreement, spanRecordsToTraceAnalysisStore, spearmanR, stripFencedJson, studentTTwoSidedPValue, subjectiveEval, summarizeBackendIntegrity, summarizeNumberSeries, summarizePreferenceMemory, summarizeSample, summaryTable, toAgentProfileJson, tokenizeDomainWords, toolSpansToTraceAnalysisStore, toolWasteView, traceContract, transientDispatchFailure, twoByTwoPValue, userQuestionsForKnowledgeGaps, validateRunRecord, verbosityBias, verifyAgentProfileCell, verifyAttestation, verifyCompletion, viteDeployRunner, weightedComposite, weightedMean, welchsTTest, wilcoxonSignedRank, wilson, withAssignedFeedbackSplit, withJudgeRetry, wranglerDeployRunner };
+export { AGENT_PROFILE_KINDS, ATTESTATION_ALGORITHM, AgentEvalError, AnalystRegistry, BOOTSTRAP_GATE_MIN_N, BackendIntegrityError, BudgetBreachError, BudgetGuard, CODING_HARNESSES, ConfigError, CostAccountingIncompleteError, CostCeilingReachedError, CostLedger, CostTracker, CrossFamilyError, DEFAULT_BOUNDED_PROCESS_TIMEOUT_MS, DEFAULT_JUDGE_GATE_POLICY, DEFAULT_RED_TEAM_CORPUS, DEFAULT_TRACE_ANALYST_BUDGETS, DEFAULT_TRACE_ANALYST_KINDS, ERROR_COUNT_PATTERNS, EquivalenceProtocolError, FAILURE_CLASSES, FAILURE_MODE_KIND_SPEC, FileSystemFeedbackTrajectoryStore, FileSystemRawProviderSink, FileSystemTraceStore, FindingsStore, HARNESS_NATIVE_MODEL, HeldOutGate, InMemoryFeedbackTrajectoryStore, InMemoryRawProviderSink, InMemoryTraceStore, JudgeError, JudgeGateRefusedError, LlmCallError, LlmResponseError, MODEL_PRICING, ModelSubstitutionError, MultiLayerVerifier, NoopRawProviderSink, NotFoundError, OUTPUT_VALUE, OtlpFileTraceStore, PairwiseSteeringOptimizer, ProductClient, PromptRegistry, REDACTION_VERSION, RunIntegrityError, RunRecordValidationError, SEMANTIC_CONCEPT_JUDGE_VERSION, TraceEmitter, UNKNOWN_MODEL, VERIFICATION_STRATEGIES, VERIFICATION_STRATEGY_SOURCES, ValidationError, acquisitionPlansForKnowledgeGaps, addCalibrationVerdicts, agentProfileCellKey, agentProfileHash, agentProfileId, aggregateJudgeVerdicts, aggregateRunScore, analystRunDigest, analystRunToFeedbackTrajectory, analystRunToReviewRequests, analyzeAntiSlop, analyzeRuns, analyzeSeries, analyzeTraces, argHash, assertCapabilityHeadroom, assertCrossFamily, assertCrossFamilyServed, assertJudgeMayGate, assertNoHiddenLeak, assertProductBenchmarkRun, assertRealBackend, assertRunCaptured, assertServedModel, assertServedModels, assertSingleBackend, assessShareSafety, assignFeedbackSplit, attest, benjaminiHochberg, blendHeldout, blockingKnowledgeEval, bonferroni, bootstrapCi, budgetBreachView, buildAgentProfileCell, buildDefaultAnalystRegistry, buildReflectionPrompt, buildTraceInsightContext, buildTraceInsightPrompt, buildTrajectory, calibrateJudge, calibrateJudgeContinuous, canonicalJson, capabilityHeadroom, captureFetchToRawSink, certificationEvidenceDigest, checkCanaries, checkServedModel, checkTraceContracts, chiSquared, chiSquaredPValue, clamp01, classifyFailure, classifyFailureReason, cliffsDelta, cohensD, comparePairedArms, compileTraceContractSpec, completionVerdict, computeFindingId, computeToolUseMetrics, confidenceInterval, contentHash, continuousAgreement, controlRunToFeedbackTrajectory, corpusInterRaterAgreement, corpusInterRaterAgreementFromJudgeScores, costForTokenPricing, costForUsage, costReceiptFromLlm, costReceiptFromLlmError, createAntiSlopJudge, createBoundedTraceAnalysisStore, createChatClient, createChatTraceEngine, createDspyRlmTraceEngine, createFeedbackTrajectory, createLlmCorrectnessChecker, createLlmReviewer, createTraceAnalyst, decidePairedPromotion, defineAgentEval, defineEquivalenceCheck, deployGateLayer, describeServedModelCheck, describeTraceInsightScope, diffFindings, diffScorecard, discoverPersonas, domainEvidencePattern, dominates, eProcess, ensembleJudge, equivalenceVerdict, errorStreakDetector, estimateCost, estimateTokens, evaluateActionPolicy, evaluateInterimReleaseConfidence, evaluateOracles, evaluateReleaseConfidence, evaluateTraceContract, expandProfileAxes, explainTraceContract, exportProductBenchmark, exportProductBenchmarkRuns, exportRunAsOtlp, extractErrorCount, extractProducedState, extractUsage, extractUsageFromSse, failureBlame, failureClusterView, feedbackTrajectoriesToDatasetScenarios, feedbackTrajectoriesToOptimizerRows, feedbackTrajectoryToOptimizerRow, fileVerdictCache, fisherExactTwoSided, formatScorecardDiff, gainHistogram, gateTreatmentApplied, gradeSemanticStatus, groupRunsByAgentProfileCell, harnessAxisOf, hashContent, hashJson, holm, inMemoryReviewStore, inferDomainKeywords, interRaterReliability, interpretCliffs, iqr, isJudgeSpan, isLlmSpan, isModelPriced, isRunRecord, isToolSpan, isTransientLlmError, jsonShape, jsonlReviewStore, jsonlRunRecordBackend, judgeAgreementView, judgeFamily, judgeGateDecision, judgeSpans, knowledgeReadinessTracePayload, leaderboard, linearRegressionSlope, lintTraceContractSpec, llmJudge, loadScorecard, localCommandRunner, makeFinding, makePolicyEditCandidateRecord, makeProposalFinding, manifestContentDigest, mannWhitneyU, maximumChargeForLlmRequest, mcnemar, mcnemarPower, mcnemarRequiredN, measureJudgeAgreement, minimumPairsForPairedDeltaTest, mintRolloutRows, modelHasSnapshot, mulberry32, normalTwoSidedPValue, objectiveEval, observeAll, otlpTextToTraceAnalysisStore, pairArms, pairRunRecords, pairedBootstrap, pairedCohensDz, pairedDeltaTest, pairedEvalueSequence, pairedMde, pairedPromotionPower, pairedRiskDifference, pairedRiskDifferenceExact, pairedRiskDifferenceScore, pairedSignTest, pairedTTest, paretoFrontier, parseReflectionResponse, parseRunRecordSafe, partialCredit, partitionHeldOut, passAtK, pearsonR, preflightModels, productBenchmarkRepoIdentity, profile_exports as profile, projectRuntimeTrajectoryEvidence, proposeSynthesisTargets, ranks, readProductBenchmarkManifest, recordRuns, recordRunsToScorecard, redTeamDataset, redTeamReport, redact, redactForShare, redactText, registerCalibrationSet, renderPreferenceMemoryMarkdown, repeatedActionDetector, requiredPairedSampleSize, requiredPairsForPairedPromotion, requiredSampleSize, resolveModelPricing, resolveSeat, runAgentControlLoop, runBoundedProcess, runCampaign, runCanaries, runEquivalenceCheck, runEvalCampaign, runIntentMatchJudge, runKeywordCoverageJudge, runKeywordCoverageJudgeUrl, runProposeReview, runSemanticConceptJudge, runTaskScore, runsForScenario, sampleMedian, scoreKnowledgeReadiness, scoreRedTeamOutput, scoreTraceInsightReadiness, seatPresets, selfImprove, shareAllowed, snapshotFromJudgeAgreement, spanRecordsToTraceAnalysisStore, spearmanR, stripFencedJson, studentTTwoSidedPValue, subjectiveEval, summarizeBackendIntegrity, summarizeNumberSeries, summarizePreferenceMemory, summarizeSample, summaryTable, toAgentProfileJson, tokenizeDomainWords, toolSpansToTraceAnalysisStore, toolWasteView, traceContract, transientDispatchFailure, twoByTwoPValue, userQuestionsForKnowledgeGaps, validateRunRecord, verbosityBias, verifyAgentProfileCell, verifyAttestation, verifyCompletion, viteDeployRunner, weightedComposite, weightedMean, welchsTTest, wilcoxonSignedRank, wilson, withAssignedFeedbackSplit, withJudgeRetry, wranglerDeployRunner };
 
 //# sourceMappingURL=index.js.map