@tangle-network/agent-eval 0.139.1 → 0.139.2

package/CHANGELOG.md

@@ -4,6 +4,13 @@ All notable changes to `@tangle-network/agent-eval` and its sibling `agent-eval-
 
 ---
 
+## [0.139.2] - 2026-07-31 - interface 0.40 cohort alignment
+
+### Fixed
+
+- Depend on `agent-interface@0.40.0`, matching the interface every current consumer ships.
+  agent-runtime's packed-cohort verifier requires this package's interface dependency to equal the packed cohort interface exactly; 0.139.1 still depended on 0.39.0.
+
 ## [0.139.1] - 2026-07-31 - source-only build gate
 
 ### Fixed

package/dist/analyst/index.d.ts

@@ -599,7 +599,7 @@ declare const ANALYST_BENCHMARK_HELP = "agent-eval analyst-benchmark\n\nRun the
 declare const ANALYST_BENCHMARK_IMPLEMENTATION_DIGEST_ALGORITHM = "sha256-canonical-source-manifest";
 declare const ANALYST_BENCHMARK_DEPENDENCY_LOCK_DIGEST_ALGORITHM = "sha256-canonical-file-manifest";
 declare const ANALYST_BENCHMARK_DEPENDENCY_LOCK_FILES: readonly string[];
-declare const ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256 = "56998607c7f9b221abeff4e96ec35a87fadd27322456a5da26f52e41133fdbbd";
+declare const ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256 = "1ca280998a9f3a416b40cadb9c14789ada490a727a4ec78491c92bb2022e6c95";
 /** The published benchmark evidence was produced at this package version, by
  * the retired one-shot direct runner, before trace analysts moved to the
  * recursive DSPy RLM engine. Both evidence digests below are historical facts

package/dist/analyst/index.js

@@ -6,7 +6,7 @@ import { n as isProposalFinding, t as assertProposalFindings } from "../proposal
 import { a as RunCritic, c as buildSkillUsageReport, d as defaultIsMaterial, f as diffFindings, h as defineTraceAnalyst, i as runSemanticConceptJudge, l as emitSkillUsageFindings, m as defineCustomAnalyst, n as SEMANTIC_CONCEPT_JUDGE_VERSION, o as SKILL_USAGE_ANALYST, s as SkillUsageAnalyst, u as FindingsStore } from "../semantic-concept-judge-BuIJ9IfB.js";
 import { t as createDspyRlmTraceEngine } from "../dspy-rlm-engine-DTkVyDX-.js";
 import { a as scoreAnalystFindings, n as runAnalystBenchmark, r as traceStoreEvidenceResolver, t as registryBenchmarkRunner } from "../benchmark-CYtcIF2V.js";
-import { A as ANALYST_BENCHMARK_IMPLEMENTATION_FILES, B as summarizeAgentRxCalibration, C as ANALYST_BENCHMARK_DEPENDENCY_LOCK_DIGEST_ALGORITHM, D as ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256, E as ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256, F as ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE, G as normalizeAgentRxCategory, H as codeTracerPredictionsToFindings, I as ANALYST_BENCHMARK_MANIFEST_FILE, K as roundAgentRxStep, L as ANALYST_BENCHMARK_OBSERVATIONS_FILE, M as analystBenchmarkDependencyLockDigest, N as analystBenchmarkImplementationDigest, O as ANALYST_BENCHMARK_EVIDENCE_PACKAGE_VERSION, P as ANALYST_BENCHMARK_COST_LEDGER_FILE, R as AGENT_RX_UPSTREAM_REVISION, S as emptyPublicBenchmarkRunner, T as ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256, U as agentRxBenchmarkCase, V as codeTraceBenchCase, W as agentRxPredictionsToFindings, _ as DEFAULT_MAX_VERIFICATION_ARTIFACT_BYTES, a as renderCodeTraceCalibrationMarkdown, b as parseVerificationOutcome, c as createPublicBenchmarkRlmRunner, d as publicBenchmarkProtocolSha256, f as loadPublicBenchmarkRows, g as selectPublicBenchmarkRows, h as publicBenchmarkSelectionReport, i as readAnalystBenchmarkArtifact, j as ANALYST_BENCHMARK_IMPLEMENTATION_SHA256, k as ANALYST_BENCHMARK_IMPLEMENTATION_DIGEST_ALGORITHM, l as CODE_TRACE_BENCH_ANALYST_PROMPT, m as publicBenchmarkDistributions, n as runAnalystBenchmarkCommand, o as summarizeCodeTraceCalibration, p as preparePublicAnalystBenchmark, q as normalizeBenchmarkLabel, r as renderAnalystBenchmarkMarkdown, s as compareAnalystRunners, t as ANALYST_BENCHMARK_HELP, u as createPublicBenchmarkDirectRunner, v as appendVerificationArtifactsToOtlp, w as ANALYST_BENCHMARK_DEPENDENCY_LOCK_FILES, x as adaptPublicBenchmarkFindings, y as loadCodeTraceVerificationArtifacts, z as renderAgentRxCalibrationMarkdown } from "../benchmark-command-Cg_PRuov.js";
+import { A as ANALYST_BENCHMARK_IMPLEMENTATION_FILES, B as summarizeAgentRxCalibration, C as ANALYST_BENCHMARK_DEPENDENCY_LOCK_DIGEST_ALGORITHM, D as ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256, E as ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256, F as ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE, G as normalizeAgentRxCategory, H as codeTracerPredictionsToFindings, I as ANALYST_BENCHMARK_MANIFEST_FILE, K as roundAgentRxStep, L as ANALYST_BENCHMARK_OBSERVATIONS_FILE, M as analystBenchmarkDependencyLockDigest, N as analystBenchmarkImplementationDigest, O as ANALYST_BENCHMARK_EVIDENCE_PACKAGE_VERSION, P as ANALYST_BENCHMARK_COST_LEDGER_FILE, R as AGENT_RX_UPSTREAM_REVISION, S as emptyPublicBenchmarkRunner, T as ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256, U as agentRxBenchmarkCase, V as codeTraceBenchCase, W as agentRxPredictionsToFindings, _ as DEFAULT_MAX_VERIFICATION_ARTIFACT_BYTES, a as renderCodeTraceCalibrationMarkdown, b as parseVerificationOutcome, c as createPublicBenchmarkRlmRunner, d as publicBenchmarkProtocolSha256, f as loadPublicBenchmarkRows, g as selectPublicBenchmarkRows, h as publicBenchmarkSelectionReport, i as readAnalystBenchmarkArtifact, j as ANALYST_BENCHMARK_IMPLEMENTATION_SHA256, k as ANALYST_BENCHMARK_IMPLEMENTATION_DIGEST_ALGORITHM, l as CODE_TRACE_BENCH_ANALYST_PROMPT, m as publicBenchmarkDistributions, n as runAnalystBenchmarkCommand, o as summarizeCodeTraceCalibration, p as preparePublicAnalystBenchmark, q as normalizeBenchmarkLabel, r as renderAnalystBenchmarkMarkdown, s as compareAnalystRunners, t as ANALYST_BENCHMARK_HELP, u as createPublicBenchmarkDirectRunner, v as appendVerificationArtifactsToOtlp, w as ANALYST_BENCHMARK_DEPENDENCY_LOCK_FILES, x as adaptPublicBenchmarkFindings, y as loadCodeTraceVerificationArtifacts, z as renderAgentRxCalibrationMarkdown } from "../benchmark-command-DzUZJl8M.js";
 //#region src/analyst/adapters.ts
 /**
 * Adapter factories — lift each existing agent-eval primitive into the

package/dist/{benchmark-command-Cg_PRuov.js → benchmark-command-DzUZJl8M.js}

@@ -1172,7 +1172,7 @@ const ANALYST_BENCHMARK_DEPENDENCY_LOCK_FILES = Object.freeze([
 	"package.json",
 	"pnpm-lock.yaml"
 ]);
-const ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256 = "56998607c7f9b221abeff4e96ec35a87fadd27322456a5da26f52e41133fdbbd";
+const ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256 = "1ca280998a9f3a416b40cadb9c14789ada490a727a4ec78491c92bb2022e6c95";
 /** The published benchmark evidence was produced at this package version, by
 * the retired one-shot direct runner, before trace analysts moved to the
 * recursive DSPy RLM engine. Both evidence digests below are historical facts
@@ -4534,4 +4534,4 @@ function shellQuote(value) {
 //#endregion
 export { ANALYST_BENCHMARK_IMPLEMENTATION_FILES as A, summarizeAgentRxCalibration as B, ANALYST_BENCHMARK_DEPENDENCY_LOCK_DIGEST_ALGORITHM as C, ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256 as D, ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256 as E, ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE as F, normalizeAgentRxCategory as G, codeTracerPredictionsToFindings as H, ANALYST_BENCHMARK_MANIFEST_FILE as I, roundAgentRxStep as K, ANALYST_BENCHMARK_OBSERVATIONS_FILE as L, analystBenchmarkDependencyLockDigest as M, analystBenchmarkImplementationDigest as N, ANALYST_BENCHMARK_EVIDENCE_PACKAGE_VERSION as O, ANALYST_BENCHMARK_COST_LEDGER_FILE as P, AGENT_RX_UPSTREAM_REVISION as R, emptyPublicBenchmarkRunner as S, ANALYST_BENCHMARK_DEPENDENCY_LOCK_SHA256 as T, agentRxBenchmarkCase as U, codeTraceBenchCase as V, agentRxPredictionsToFindings as W, DEFAULT_MAX_VERIFICATION_ARTIFACT_BYTES as _, renderCodeTraceCalibrationMarkdown as a, parseVerificationOutcome as b, createPublicBenchmarkRlmRunner as c, publicBenchmarkProtocolSha256 as d, loadPublicBenchmarkRows as f, selectPublicBenchmarkRows as g, publicBenchmarkSelectionReport as h, readAnalystBenchmarkArtifact as i, ANALYST_BENCHMARK_IMPLEMENTATION_SHA256 as j, ANALYST_BENCHMARK_IMPLEMENTATION_DIGEST_ALGORITHM as k, CODE_TRACE_BENCH_ANALYST_PROMPT as l, publicBenchmarkDistributions as m, runAnalystBenchmarkCommand as n, summarizeCodeTraceCalibration as o, preparePublicAnalystBenchmark as p, normalizeBenchmarkLabel as q, renderAnalystBenchmarkMarkdown as r, compareAnalystRunners as s, ANALYST_BENCHMARK_HELP as t, createPublicBenchmarkDirectRunner as u, appendVerificationArtifactsToOtlp as v, ANALYST_BENCHMARK_DEPENDENCY_LOCK_FILES as w, adaptPublicBenchmarkFindings as x, loadCodeTraceVerificationArtifacts as y, renderAgentRxCalibrationMarkdown as z };
 
-//# sourceMappingURL=benchmark-command-Cg_PRuov.js.map
+//# sourceMappingURL=benchmark-command-DzUZJl8M.js.map