@tangle-network/agent-eval 0.120.4 → 0.121.0

package/dist/fuzz.js

@@ -1,7 +1,7 @@
 import {
   CostLedger,
   CostReceiptCaptureError
-} from "./chunk-ZMXDQ4K7.js";
+} from "./chunk-JHOJHHU7.js";
 import "./chunk-VI2UW6B6.js";
 import {
   varianceBasedCurriculum

package/dist/index.js

@@ -9,12 +9,12 @@ import {
   checkBehavioralCanary,
   checkCanaries,
   runBehavioralCanaries
-} from "./chunk-SQQED7ZH.js";
+} from "./chunk-3A246TSA.js";
 import {
   BENCHMARK_SPLIT_SEED,
   benchmarks_exports,
   deterministicSplit
-} from "./chunk-CVJP5TMD.js";
+} from "./chunk-PICTDURQ.js";
 import {
   DEFAULT_RULES,
   buildTrajectory,
@@ -51,7 +51,7 @@ import {
   harnessAxisOf,
   parseCorrectnessResponse,
   verifyCompletion
-} from "./chunk-ZTJPIIVI.js";
+} from "./chunk-JN2FCO5W.js";
 import {
   DEFAULT_MUTATION_PRIMITIVES,
   DEFAULT_RED_TEAM_CORPUS,
@@ -91,7 +91,7 @@ import {
   scoreRedTeamOutput,
   surfaceContentHash,
   toolNamesForRun
-} from "./chunk-7QBFOJ3E.js";
+} from "./chunk-32BZXMSO.js";
 import {
   BackendIntegrityError,
   assertRealAgentReceipts,
@@ -103,7 +103,7 @@ import {
   inMemoryVerdictCache,
   summarizeAgentReceiptIntegrity,
   summarizeBackendIntegrity
-} from "./chunk-SYV364BL.js";
+} from "./chunk-XDIRG3TO.js";
 import {
   DEFAULT_COMPLEXITY_WEIGHTS,
   FindingsStore,
@@ -116,12 +116,12 @@ import {
   defaultIsMaterial,
   diffFindings,
   runSemanticConceptJudge
-} from "./chunk-3XH4Y2SS.js";
+} from "./chunk-JM2SKQMS.js";
 import {
   buildDefaultAnalystRegistry,
   computeTraceMetrics,
   createChatClient
-} from "./chunk-5BYTIDZ7.js";
+} from "./chunk-QWMPPZ3X.js";
 import "./chunk-HHWE3POT.js";
 import {
   Mutex
@@ -157,7 +157,7 @@ import {
   scorePolicyEditReadiness,
   validatePolicyEdit,
   validatePolicyEditCandidateRecord
-} from "./chunk-5CVUPHJ4.js";
+} from "./chunk-S5TT5R3L.js";
 import {
   allCriticalPassed,
   controlFailureClassFromVerification,
@@ -267,7 +267,7 @@ import {
   CostReservationExceededError,
   costForUsage,
   modelPriceKey
-} from "./chunk-ZMXDQ4K7.js";
+} from "./chunk-JHOJHHU7.js";
 import {
   MODEL_PRICING,
   MetricsCollector,

package/dist/openapi.json

@@ -2,7 +2,7 @@
   "openapi": "3.1.0",
   "info": {
     "title": "@tangle-network/agent-eval — wire protocol",
-    "version": "0.120.4",
+    "version": "0.121.0",
     "description": "HTTP and stdio RPC interface to agent-eval. The TypeScript runtime is the source of truth; this spec is the contract that cross-language clients (Python, Rust, Go) generate from.\n\nWire-protocol version: 1.0.0. Bumps on breaking changes to request/response schemas.",
     "contact": {
       "name": "Tangle Network",

package/dist/{run-campaign-75RTPVV5.js → run-campaign-HNFPJET4.js}

@@ -1,9 +1,9 @@
 import {
   planCampaignRun,
   runCampaign
-} from "./chunk-SYV364BL.js";
+} from "./chunk-XDIRG3TO.js";
 import "./chunk-PJQFMIOX.js";
-import "./chunk-ZMXDQ4K7.js";
+import "./chunk-JHOJHHU7.js";
 import "./chunk-VI2UW6B6.js";
 import "./chunk-ONWEPEDO.js";
 import "./chunk-PZ5AY32C.js";
@@ -11,4 +11,4 @@ export {
   planCampaignRun,
   runCampaign
 };
-//# sourceMappingURL=run-campaign-75RTPVV5.js.map
+//# sourceMappingURL=run-campaign-HNFPJET4.js.map

package/dist/wire/index.js

@@ -34,9 +34,9 @@ import {
   runRpcOnce,
   startServer,
   startServerAsync
-} from "../chunk-NMN4WGSJ.js";
+} from "../chunk-4I2E3LLO.js";
 import "../chunk-NJC7U437.js";
-import "../chunk-ZMXDQ4K7.js";
+import "../chunk-JHOJHHU7.js";
 import "../chunk-VI2UW6B6.js";
 import "../chunk-PC4UYEBM.js";
 import "../chunk-ONWEPEDO.js";

package/docs/eval-surface-map.md

@@ -12,7 +12,7 @@ composition point already exists (see Produced-state grading below).**
 | `runCampaign` | The measurement primitive. Run a dispatch over scenarios × seeds × reps, score each with judges, aggregate. Caller owns the dispatch. | `CampaignResult` |
 | `runEval` | The simplest preset over `runCampaign` — just score, no loop, no gate. The 80% "I want a scorecard" case. | `CampaignResult` |
 | `runProfileMatrix` | Factor the SAME scenarios across N agent **profiles** (models / prompt variants), with RunRecord stamping + a real-backend integrity guard. | `RunRecord[]` |
-| `runOptimization` | GENERATE: measured or validated premeasured baseline → N generations of propose → measure → rank → promote. No release gate. | generations + winner |
+| `runOptimization` | GENERATE: baseline → N generations of propose → measure → rank → promote. No release gate. | generations + winner |
 | `runImprovementLoop` | The release-gate shell around `runOptimization`: adds a held-out re-score + a promotion gate (+ optional auto-PR). | gate decision + winner |
 | `runEvalCampaign` | Inversion-of-control variant of `runCampaign` — the runner is handed a pre-wired trace/sink/emitter and integrity gating as a precondition. Use when you need full capture by construction. | `CampaignResult` + records |
 | `runAgentMatrix` | The bare N-axis cartesian scheduler with concurrency control. The layer beneath the eval surface — reach for it only when you need raw scheduling, not eval semantics. | cell results |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tangle-network/agent-eval",
-  "version": "0.120.4",
+  "version": "0.121.0",
   "description": "Evaluate and improve AI agents from runs, traces, judges, and feedback. Compare candidates, cluster failures, measure lift, and gate releases.",
   "homepage": "https://github.com/tangle-network/agent-eval#readme",
   "repository": {
@@ -152,7 +152,7 @@
     "@ax-llm/ax": "^23.0.0",
     "@hono/node-server": "^2.0.0",
     "@tangle-network/agent-core": "^0.4.11",
-    "@tangle-network/agent-interface": "^0.26.0",
+    "@tangle-network/agent-interface": "^0.28.0",
     "@tangle-network/tcloud": "^0.4.14",
     "hono": "^4.12.30",
     "zod": "^4.3.6"