@tangle-network/agent-eval 1.0.7 → 1.0.8

package/CHANGELOG.md

@@ -4,6 +4,26 @@ All notable changes to `@tangle-network/agent-eval` and its sibling `agent-eval-
 
 ---
 
+## [1.0.8] — 2026-10-30
+
+- chore(analyst): re-pin the implementation manifest on merged main
+- feat(statistics,harness): two-sample Welch evidence + lifted sandbox assertion library
+- chore(release): 1.0.7 (#1172)
+- docs(statistics): keep inference policy and exports readable (#1173)
+- feat(statistics): own independent effect summaries and ROC area (#1171)
+- chore(release): 1.0.6 (#1170)
+- refactor(reporting): share text table layout and simplify summaries (#1169)
+- refactor(harness): retain unavailable verdicts through cleanup (#1168)
+- chore(release): 1.0.5 (#1167)
+- refactor(harness): normalize compact named expectations (#1166)
+- chore(release): 1.0.4 (#1165)
+- refactor(harness): own resource acquisition and readiness cleanup (#1164)
+- chore(release): 1.0.3 (#1163)
+- chore(release): declare the existing deployment runner for workflow lint
+- fix(release): bind bootstrap to the existing qualified action revision
+- fix(release): use the qualified pnpm bootstrap in publication workflows
+- Refactor sandbox integration scenarios into the eval harness (#1161)
+
 ## [1.0.7] — 2026-10-29
 
 - feat(statistics): own independent effect summaries and ROC area (#1171)

package/dist/analyst/index.js

@@ -11,7 +11,7 @@ import { t as createDspyRlmTraceEngine } from "../dspy-rlm-engine-Xt3znknf.js";
 import { n as diffFindings, r as createChatTraceEngine, t as FindingsStore } from "../findings-store-Bp-0Guuv.js";
 import { a as systemOneAnalyst, p as validDecisionIdentity } from "../decision-point-CErE7GnP.js";
 import { t as createChatClient } from "../chat-client-D0me8iwq.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-BO-YeQaH.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-CyNmaPml.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-hNXESoc1.js";

package/dist/{benchmark-command-BO-YeQaH.js → benchmark-command-CyNmaPml.js}

@@ -1998,6 +1998,7 @@ const ANALYST_BENCHMARK_IMPLEMENTATION_FILES = Object.freeze([
 	"src/statistics/rank-tests.ts",
 	"src/statistics/samples.ts",
 	"src/statistics/sequential-eprocess.ts",
+	"src/statistics/two-sample.ts",
 	"src/trace-analyst/errors.ts",
 	"src/trace-analyst/otlp-span.ts",
 	"src/trace-analyst/shared-abortable-task.ts",
@@ -2019,7 +2020,7 @@ const ANALYST_BENCHMARK_IMPLEMENTATION_FILES = Object.freeze([
 	"src/trace/redaction/detectors.ts",
 	"src/verdict-cache.ts"
 ]);
-const ANALYST_BENCHMARK_IMPLEMENTATION_SHA256 = "dc3a548100175edf1a072c2f287c8ccaf8eadde01be1dd54a6e09a6181dfc7b6";
+const ANALYST_BENCHMARK_IMPLEMENTATION_SHA256 = "e58b68ae8112794a8b1e615b6372a5d2a145e4374cee7ff3590ce2c630c26cc1";
 // src/analyst/benchmark-evidence-validation.ts
 const MIN_ACTION_EXCERPT_CHARACTERS = 12;
 const MAX_LABEL_SCAN_DEPTH = 64;
@@ -6856,4 +6857,4 @@ function shellQuote(value) {
 }
 export { agentRxBenchmarkCase as A, ANALYST_BENCHMARK_COST_LEDGER_FILE as C, AGENT_RX_UPSTREAM_REVISION as D, ANALYST_BENCHMARK_OBSERVATIONS_FILE as E, runAnalystBenchmark as F, traceStoreEvidenceResolver as I, normalizeAgentRxCategory as M, externalId as N, codeTraceBenchCase as O, positiveStep as P, ANALYST_BENCHMARK_IMPLEMENTATION_FILES as S, ANALYST_BENCHMARK_MANIFEST_FILE as T, readAnalystBenchmarkArtifact as _, runInlineAnalystDefinition as a, ANALYST_BENCHMARK_EVIDENCE_DEPENDENCY_LOCK_SHA256 as b, publicRlmAnalystDefinition as c, publicDirectAnalystDefinition as d, runChunkedAnalystDefinition as f, expandCodeTraceFailureBlocks as g, analystDefinitionProtocolSha256 as h, primeCodeTraceAnalystDefinition as i, agentRxPredictionsToFindings as j, codeTracerPredictionsToFindings as k, runReplVariableAnalystDefinition as l, analystDefinitionAsymmetries as m, createPrimeBenchmarkRunner as n, loadPublicBenchmarkRows as o, AnalystExpressivenessError as p, primeAnalystProtocolSha256 as r, createPublicBenchmarkRlmRunner as s, runAnalystBenchmarkCommand as t, createPublicBenchmarkDirectRunner as u, CODE_TRACE_BENCH_ANALYST_PROMPT as v, ANALYST_BENCHMARK_LOCAL_RECEIPT_FILE as w, ANALYST_BENCHMARK_EVIDENCE_IMPLEMENTATION_SHA256 as x, publicBenchmarkProtocolSha256 as y };
 
-//# sourceMappingURL=benchmark-command-BO-YeQaH.js.map
+//# sourceMappingURL=benchmark-command-CyNmaPml.js.map