@tangle-network/agent-eval 0.221.8 → 0.222.0

package/dist/analyst/index.js

@@ -1,25 +1,25 @@
-import { n as assertConsistentFindingConditions, o as mergeFindingComposition, r as findingCompositionKey, t as TREATMENT_LAYERS } from "../finding-composition-8n0fFgVI.js";
-import { i as nominalAgreement } from "../agreement-irr-BhAqjluA.js";
-import { t as FileLedgerJournal } from "../journal-C7om8Gum.js";
-import { r as merkleRoot } from "../ledger-core-CAxUdDk0.js";
-import { A as spanningWindow, C as UNKNOWN_LINEAGE, D as mostSevere, E as mergeEvidenceUnits, O as parseAnalystFinding, S as ProposalFindingSchema, T as lineageRoots, b as AnalystFindingSchema, k as readFindingScope, w as independentUnitCount, x as FINDING_LAYERS } from "../external-optimizer-subprocess-B_ukPtRD.js";
-import { a as makeProposalFinding, i as makeFinding, r as computeFindingId } from "../types-CnWmj8NW.js";
-import { t as assertProposalFindings } from "../proposal-findings-DqLUWsAG.js";
-import { A as RawAnalystFindingSchema, B as parseFindingSubject, F as stripCodeFences, I as FINDING_SUBJECT_KINDS, L as FINDING_SUBJECT_SYNTAX, M as parseRawFinding, N as parseTraceSpanEvidenceUri, R as KIND_EXPECTED_SUBJECTS, V as renderFindingSubject, j as evidenceRefsFromRawFinding, k as RAW_FINDING_SCHEMA_PROMPT, n as runTraceAnalyst, r as buildTraceToolsForGroup, t as createTraceAnalyst, z as findingSubjectGrammarPromptFor } from "../kind-factory-Dl0PmNQL.js";
-import { _ as behavioralTraceQuestionDigest, a as KNOWLEDGE_POISONING_KIND_SPEC, c as FAILURE_MODE_KIND_SPEC, d as emitControlIntegrityFindings, f as behavioralAnalyst, g as BEHAVIORAL_TRACE_QUESTIONS_VERSION, h as BEHAVIORAL_TRACE_QUESTIONS, i as DEFAULT_TRACE_ANALYST_KINDS, l as CONTROL_INTEGRITY_ANALYST, m as behavioralTraceSupportSchema, n as AnalystRegistry, o as KNOWLEDGE_GAP_KIND_SPEC, p as validBehavioralCalibration, r as ExactAnalystRunExecutionError, s as IMPROVEMENT_KIND_SPEC, t as buildDefaultAnalystRegistry, u as ControlIntegrityAnalyst } from "../default-registry-BBTDgkDM.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 systemOneAnalyst, p as validDecisionIdentity } from "../decision-point-0nJubuoM.js";
-import { t as createChatClient } from "../chat-client-BNATB3rJ.js";
-import { A as agentRxBenchmarkCase, C as ANALYST_BENCHMARK_COST_LEDGER_FILE, D as AGENT_RX_UPSTREAM_REVISION, E as ANALYST_BENCHMARK_OBSERVATIONS_FILE, F as runAnalystBenchmark, I as traceStoreEvidenceResolver, M as normalizeAgentRxCategory, N as externalId, O as codeTraceBenchCase, P as positiveStep, S as ANALYST_BENCHMARK_IMPLEMENTATION_FILES, T as ANALYST_BENCHMARK_MANIFEST_FILE, _ as readAnalystBenchmarkArtifact, a as runInlineAnalystDefinition, b as ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256, c as publicRlmAnalystDefinition, d as publicDirectAnalystDefinition, f as runChunkedAnalystDefinition, g as expandCodeTraceFailureBlocks, h as analystDefinitionProtocolSha256, i as primeCodeTraceAnalystDefinition, j as agentRxPredictionsToFindings, k as codeTracerPredictionsToFindings, l as runReplVariableAnalystDefinition, m as analystDefinitionAsymmetries, n as createPrimeBenchmarkRunner, o as loadPublicBenchmarkRows, p as AnalystExpressivenessError, r as primeAnalystProtocolSha256, s as createPublicBenchmarkRlmRunner, t as runAnalystBenchmarkCommand, u as createPublicBenchmarkDirectRunner, v as CODE_TRACE_BENCH_ANALYST_PROMPT, w as ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE, x as ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256, y as publicBenchmarkProtocolSha256 } from "../benchmark-command-Cnjfo0Zh.js";
-import { a as normalizePrimeUsage, c as runPrimeExchange, i as mergePrimeRawUsage, n as buildPrimePrompt, o as primeProtocolSha256, r as buildPrimeRepairPrompt, s as projectPrimeTrajectory, t as analystUsageReceiptFromPrimeUsage } from "../prime-protocol-BFgGrEMP.js";
-import { t as nodeHttpPrimeBridgeTransport } from "../prime-bridge-transport-Z3cpKGoV.js";
-import { t as clusterBootstrapMean } from "../cluster-bootstrap-DEwwhl0C.js";
+import { n as assertConsistentFindingConditions, o as mergeFindingComposition, r as findingCompositionKey, t as TREATMENT_LAYERS } from "../finding-composition-DlvE1Qul.js";
+import { i as nominalAgreement } from "../agreement-irr-BTjDN0i6.js";
+import { t as FileLedgerJournal } from "../journal-C1uLJi0w.js";
+import { r as merkleRoot } from "../ledger-core-CfT-FzPf.js";
+import { A as spanningWindow, C as UNKNOWN_LINEAGE, D as mostSevere, E as mergeEvidenceUnits, O as parseAnalystFinding, S as ProposalFindingSchema, T as lineageRoots, b as AnalystFindingSchema, k as readFindingScope, w as independentUnitCount, x as FINDING_LAYERS } from "../external-optimizer-subprocess-U616hV_l.js";
+import { a as makeProposalFinding, i as makeFinding, r as computeFindingId } from "../types-BK0blWhe.js";
+import { t as assertProposalFindings } from "../proposal-findings-WFQrxnfe.js";
+import { A as RawAnalystFindingSchema, B as parseFindingSubject, F as stripCodeFences, I as FINDING_SUBJECT_KINDS, L as FINDING_SUBJECT_SYNTAX, M as parseRawFinding, N as parseTraceSpanEvidenceUri, R as KIND_EXPECTED_SUBJECTS, V as renderFindingSubject, j as evidenceRefsFromRawFinding, k as RAW_FINDING_SCHEMA_PROMPT, n as runTraceAnalyst, r as buildTraceToolsForGroup, t as createTraceAnalyst, z as findingSubjectGrammarPromptFor } from "../kind-factory-VKysp1ds.js";
+import { _ as behavioralTraceQuestionDigest, a as KNOWLEDGE_POISONING_KIND_SPEC, c as FAILURE_MODE_KIND_SPEC, d as emitControlIntegrityFindings, f as behavioralAnalyst, g as BEHAVIORAL_TRACE_QUESTIONS_VERSION, h as BEHAVIORAL_TRACE_QUESTIONS, i as DEFAULT_TRACE_ANALYST_KINDS, l as CONTROL_INTEGRITY_ANALYST, m as behavioralTraceSupportSchema, n as AnalystRegistry, o as KNOWLEDGE_GAP_KIND_SPEC, p as validBehavioralCalibration, r as ExactAnalystRunExecutionError, s as IMPROVEMENT_KIND_SPEC, t as buildDefaultAnalystRegistry, u as ControlIntegrityAnalyst } from "../default-registry-C1tPEjZ4.js";
+import { t as createDspyRlmTraceEngine } from "../dspy-rlm-engine-jDiWkul6.js";
+import { n as diffFindings, r as createChatTraceEngine, t as FindingsStore } from "../findings-store-B75Dspaq.js";
+import { a as systemOneAnalyst, p as validDecisionIdentity } from "../decision-point-lhMgODOR.js";
+import { t as createChatClient } from "../chat-client-CO3uB33i.js";
+import { A as agentRxBenchmarkCase, C as ANALYST_BENCHMARK_COST_LEDGER_FILE, D as AGENT_RX_UPSTREAM_REVISION, E as ANALYST_BENCHMARK_OBSERVATIONS_FILE, F as runAnalystBenchmark, I as traceStoreEvidenceResolver, M as normalizeAgentRxCategory, N as externalId, O as codeTraceBenchCase, P as positiveStep, S as ANALYST_BENCHMARK_IMPLEMENTATION_FILES, T as ANALYST_BENCHMARK_MANIFEST_FILE, _ as readAnalystBenchmarkArtifact, a as runInlineAnalystDefinition, b as ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256, c as publicRlmAnalystDefinition, d as publicDirectAnalystDefinition, f as runChunkedAnalystDefinition, g as expandCodeTraceFailureBlocks, h as analystDefinitionProtocolSha256, i as primeCodeTraceAnalystDefinition, j as agentRxPredictionsToFindings, k as codeTracerPredictionsToFindings, l as runReplVariableAnalystDefinition, m as analystDefinitionAsymmetries, n as createPrimeBenchmarkRunner, o as loadPublicBenchmarkRows, p as AnalystExpressivenessError, r as primeAnalystProtocolSha256, s as createPublicBenchmarkRlmRunner, t as runAnalystBenchmarkCommand, u as createPublicBenchmarkDirectRunner, v as CODE_TRACE_BENCH_ANALYST_PROMPT, w as ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE, x as ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256, y as publicBenchmarkProtocolSha256 } from "../benchmark-command-jYWR2gya.js";
+import { a as normalizePrimeUsage, c as runPrimeExchange, i as mergePrimeRawUsage, n as buildPrimePrompt, o as primeProtocolSha256, r as buildPrimeRepairPrompt, s as projectPrimeTrajectory, t as analystUsageReceiptFromPrimeUsage } from "../prime-protocol-CrhpXRz9.js";
+import { t as nodeHttpPrimeBridgeTransport } from "../prime-bridge-transport-Csd5SQg7.js";
+import { t as clusterBootstrapMean } from "../cluster-bootstrap-SbaeM44s.js";
 import { z } from "zod";
 import { existsSync, readFileSync, readdirSync, statSync } from "node:fs";
 import { join } from "node:path";
 import { readFile } from "node:fs/promises";
-//#region src/analyst/benchmark-behavioral-calibration.ts
+// src/analyst/benchmark-behavioral-calibration.ts
 const identity$1 = z.string().refine(validDecisionIdentity);
 const questionId = z.enum(Object.keys(BEHAVIORAL_TRACE_QUESTIONS));
 const answer = z.enum([
@@ -150,8 +150,7 @@ function calibrateBehavioralTraceQuestions(input) {
 		return row;
 	});
 }
-//#endregion
-//#region src/analyst/bind.ts
+// src/analyst/bind.ts
 /**
 * Compile a definition plus a transport binding into a runnable arm. Dispatch
 * is total over the expressible pairs; everything else is a loud refusal.
@@ -168,8 +167,7 @@ function bindAnalyst(definition, transports) {
 	if (mode === "repl-variable" && transports.kind === "model-owner") return runReplVariableAnalystDefinition(definition, transports.config);
 	throw new AnalystExpressivenessError(`no strategy compiles a '${mode}' projection over a '${transports.kind}' transport; definition '${definition.id}' cannot be expressed by this layer yet`);
 }
-//#endregion
-//#region src/analyst/define.ts
+// src/analyst/define.ts
 /**
 * Define a reusable trace-research question.
 *
@@ -205,8 +203,7 @@ function defineCustomAnalyst(options) {
 		..."executionConfig" in options ? { executionConfig: options.executionConfig } : {}
 	};
 }
-//#endregion
-//#region src/analyst/domain-taxonomy.ts
+// src/analyst/domain-taxonomy.ts
 /**
 * Domain taxonomy — one versioned tree that both the layer ladder and the
 * domain hierarchy live in.
@@ -381,8 +378,7 @@ function lowestCommonAncestor(taxonomy, ids) {
 function nodeIndex(taxonomy) {
 	return new Map(taxonomy.nodes.map((node) => [node.id, node]));
 }
-//#endregion
-//#region src/analyst/failure-modes.ts
+// src/analyst/failure-modes.ts
 /**
 * failure-modes — a judged analyst with fixed-choice labels, gated on its own
 * calibration.
@@ -727,8 +723,7 @@ function createFailureModesAnalyst(options) {
 		})
 	};
 }
-//#endregion
-//#region src/analyst/findings-ledger.ts
+// src/analyst/findings-ledger.ts
 /**
 * Findings ledger — analyst findings in the shared ledger-core journal.
 *
@@ -906,8 +901,7 @@ function diffFindingsOverTime(rows, policy = {}) {
 		diff: diffFindings(byRun.get(runs[index]), byRun.get(runId), policy)
 	}));
 }
-//#endregion
-//#region src/analyst/kinds/skill-usage.ts
+// src/analyst/kinds/skill-usage.ts
 /**
 * Skill-usage analyst — a DETERMINISTIC `Analyst` over a Claude/Codex skill
 * library + its trace corpus. Unlike the trace-store kinds (failure-mode,
@@ -1056,8 +1050,7 @@ function buildSkillUsageReport(config) {
 		records
 	};
 }
-//#endregion
-//#region src/analyst/promote.ts
+// src/analyst/promote.ts
 /**
 * Promotion — the only way a claim moves above the layer it was measured at.
 *
@@ -1304,8 +1297,7 @@ function liftedFinding(voters, parent, taxonomy, producedAt) {
 		...producedAt === void 0 ? {} : { produced_at: producedAt }
 	});
 }
-//#endregion
-//#region src/analyst/rollup.ts
+// src/analyst/rollup.ts
 /**
 * rollup — run findings aggregated up the ladder, with bootstrap intervals and
 * denominators, lifted only where `promote` allows.
@@ -1475,8 +1467,7 @@ function pool(members, options, confidence) {
 		}
 	};
 }
-//#endregion
-//#region src/analyst/team-shape.ts
+// src/analyst/team-shape.ts
 const TEAM_SHAPE_ANALYST_ID = "team-shape";
 const TEAM_SHAPE_THRESHOLDS = {
 	/** Share of the root's children that delegate at or above which depth ≥ 2 is a hive. */
@@ -1915,8 +1906,7 @@ function teamShapeAnalyst() {
 		}
 	};
 }
-//#endregion
-//#region src/analyst/trace-questions.ts
+// src/analyst/trace-questions.ts
 const QUESTION_ID = /^[A-Za-z0-9][A-Za-z0-9._-]{0,63}$/u;
 const DEFINITION_VERSION = "1.0.0";
 const QUESTION_RULES = [
@@ -2120,7 +2110,6 @@ function failed(question, failure) {
 function errorText(error) {
 	return error instanceof Error ? `${error.name}: ${error.message}` : String(error);
 }
-//#endregion
 export { AGENT_RX_UPSTREAM_REVISION, ANALYST_BENCHMARK_COST_LEDGER_FILE, ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256, ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256, ANALYST_BENCHMARK_IMPLEMENTATION_FILES, ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE, ANALYST_BENCHMARK_MANIFEST_FILE, ANALYST_BENCHMARK_OBSERVATIONS_FILE, AnalystExpressivenessError, AnalystRegistry, BEHAVIORAL_TRACE_QUESTIONS, BEHAVIORAL_TRACE_QUESTIONS_VERSION, CODE_TRACE_BENCH_ANALYST_PROMPT, CONTROL_INTEGRITY_ANALYST, ControlIntegrityAnalyst, DEFAULT_TRACE_ANALYST_KINDS, ExactAnalystRunExecutionError, FAILURE_MODES_ANALYST_ID, FAILURE_MODES_MIN_KAPPA, FAILURE_MODES_QUESTIONS, FAILURE_MODES_VERSION, FAILURE_MODE_KIND_SPEC, FAILURE_MODE_LABELS, FINDING_LAYERS, FINDING_SUBJECT_KINDS, FINDING_SUBJECT_SYNTAX, FindingsStore, IMPROVEMENT_KIND_SPEC, KIND_EXPECTED_SUBJECTS, KNOWLEDGE_GAP_KIND_SPEC, KNOWLEDGE_POISONING_KIND_SPEC, LAB_NODE_ID, NOT_OBSERVABLE, ProposalFindingSchema, RAW_FINDING_SCHEMA_PROMPT, RawAnalystFindingSchema, TEAM_SHAPE_ANALYST_ID, TEAM_SHAPE_CLASSES, TEAM_SHAPE_VERSION, TREATMENT_LAYERS, agentRxBenchmarkCase, agentRxPredictionsToFindings, analystDefinitionAsymmetries, analystDefinitionProtocolSha256, analystUsageReceiptFromPrimeUsage, askTraceQuestions, assertProposalFindings, attachToTaxonomy, behavioralAnalyst, behavioralTraceQuestionDigest, bindAnalyst, buildDefaultAnalystRegistry, buildPrimePrompt, buildPrimeRepairPrompt, buildSkillUsageReport, buildTraceToolsForGroup, calibrateBehavioralTraceQuestions, calibrateFailureModes, classifyTeamShape, codeTraceBenchCase, codeTracerPredictionsToFindings, computeFindingId, createChatClient, createChatTraceEngine, createDspyRlmTraceEngine, createFailureModesAnalyst, createPrimeBenchmarkRunner, createPublicBenchmarkDirectRunner, createPublicBenchmarkRlmRunner, createTraceAnalyst, defineCustomAnalyst, defineDomainTaxonomy, defineTraceAnalyst, diffFindings, diffFindingsOverTime, domainPathOf, emitControlIntegrityFindings, evidenceRefsFromRawFinding, expandCodeTraceFailureBlocks, findingSubjectGrammarPromptFor, independentUnitCount, lineageRoots, loadFailureModeGold, loadPublicBenchmarkRows, lowestCommonAncestor, makeFinding, makeProposalFinding, mergePrimeRawUsage, nodeHttpPrimeBridgeTransport, normalizePrimeUsage, openFindingsLedger, parseAnalystFinding, parseFindingSubject, parseRawFinding, primeAnalystProtocolSha256, primeCodeTraceAnalystDefinition, primeProtocolSha256, projectPrimeTrajectory, promote, publicBenchmarkProtocolSha256, publicDirectAnalystDefinition, publicRlmAnalystDefinition, queryFindings, readAnalystBenchmarkArtifact, readFindingScope, renderFindingSubject, resolveDomainPath, rollupFindings, runAnalystBenchmark, runAnalystBenchmarkCommand, runPrimeExchange, runTraceAnalyst, stripCodeFences, taxonomyAncestry, taxonomyLayerOf, teamShapeAnalyst, teamShapeInputFromSources, traceStoreEvidenceResolver };
 
 //# sourceMappingURL=index.js.map