@tangle-network/agent-eval 0.101.1 → 0.102.0

package/dist/campaign/index.d.ts

@@ -3,9 +3,9 @@ export { L as LlmJudgeDimension, c as LlmJudgeOptions, l as llmJudge } from '../
 import { A as AnalyzeTracesOptions, a as AnalyzeTracesInput, b as AnalyzeTracesResult } from '../analyst-C8HHvfJp.js';
 import { S as Scenario, M as MutableSurface, b as DispatchContext, a as JudgeConfig, e as GenerationRecord, g as Gate, J as JudgeScore, L as LabeledScenarioStore, s as LabeledScenarioWrite, t as LabeledScenarioSampleArgs, u as LabeledScenarioRecord, v as LabelTrust, f as SurfaceProposer, w as ProposedCandidate, x as ProposeContext, y as LabeledScenarioSource, C as CampaignResult, o as CodeSurface } from '../types-fWqEJm7h.js';
 export { k as CampaignAggregates, l as CampaignArtifactWriter, m as CampaignCellResult, n as CampaignCostMeter, z as CampaignTokenUsage, d as CampaignTraceWriter, D as DispatchFn, h as GateContext, j as GateDecision, G as GateResult, p as GenerationCandidate, A as JudgeAggregate, c as JudgeDimension, i as Mutator, O as OptimizationProposer, q as OptimizerConfig, P as ParetoParent, R as RedactionStatus, B as ScenarioAggregate, r as SessionScript, T as TraceSpan, E as isProposedCandidate, F as labelTrustRank } from '../types-fWqEJm7h.js';
-import { e as CampaignRunPlan, P as PlanCampaignRunOptions, C as CampaignStorage, R as RunCampaignOptions, c as RunImprovementLoopOptions } from '../gepa-CEy1AIWp.js';
-export { h as CampaignRunPlanCell, j as GepaProposerConstraints, G as GepaProposerOptions, O as OpenAutoPrOptions, k as OpenAutoPrResult, b as RunImprovementLoopResult, a as RunOptimizationOptions, l as RunOptimizationResult, m as countSentenceEdits, n as defaultRenderDiff, o as extractH2Sections, f as fsCampaignStorage, g as gepaProposer, i as inMemoryCampaignStorage, p as openAutoPr, q as planCampaignRun, r as runCampaign, d as runImprovementLoop, s as runOptimization, t as surfaceHash } from '../gepa-CEy1AIWp.js';
-export { A as AxisEvidence, a as AxisVerdict, B as BuildEvidenceVectorOptions, k as BuildLoopProvenanceArgs, D as DefaultProductionGateOptions, l as EmitLoopProvenanceArgs, m as EmitLoopProvenanceResult, E as EvidenceVector, b as EvolutionaryProposerOptions, H as HeldOutGateOptions, n as LoopProvenanceBackend, o as LoopProvenanceCandidate, L as LoopProvenanceRecord, O as ObjectiveSource, P as ParetoSignificanceGateOptions, c as PromotionObjective, d as PromotionPolicy, R as RunEvalOptions, e as buildEvidenceVector, q as buildLoopProvenanceRecord, f as composeGate, g as defaultProductionGate, s as emitLoopProvenance, h as evolutionaryProposer, i as heldOutGate, t as loopProvenanceSpans, p as paretoPolicy, j as paretoSignificanceGate, u as provenanceRecordPath, v as provenanceSpansPath, r as runEval, w as surfaceContentHash } from '../provenance-DdfmVfqR.js';
+import { e as CampaignRunPlan, P as PlanCampaignRunOptions, C as CampaignStorage, R as RunCampaignOptions, c as RunImprovementLoopOptions } from '../gepa-bxuDoaO9.js';
+export { h as CampaignRunPlanCell, j as GepaProposerConstraints, G as GepaProposerOptions, O as OpenAutoPrOptions, k as OpenAutoPrResult, b as RunImprovementLoopResult, a as RunOptimizationOptions, l as RunOptimizationResult, m as countSentenceEdits, n as defaultRenderDiff, o as extractH2Sections, f as fsCampaignStorage, g as gepaProposer, i as inMemoryCampaignStorage, p as openAutoPr, q as planCampaignRun, r as runCampaign, d as runImprovementLoop, s as runOptimization, t as surfaceHash } from '../gepa-bxuDoaO9.js';
+export { A as AxisEvidence, a as AxisVerdict, B as BuildEvidenceVectorOptions, k as BuildLoopProvenanceArgs, D as DefaultProductionGateOptions, l as EmitLoopProvenanceArgs, m as EmitLoopProvenanceResult, E as EvidenceVector, b as EvolutionaryProposerOptions, H as HeldOutGateOptions, n as LoopProvenanceBackend, o as LoopProvenanceCandidate, L as LoopProvenanceRecord, O as ObjectiveSource, P as ParetoSignificanceGateOptions, c as PromotionObjective, d as PromotionPolicy, R as RunEvalOptions, e as buildEvidenceVector, q as buildLoopProvenanceRecord, f as composeGate, g as defaultProductionGate, s as emitLoopProvenance, h as evolutionaryProposer, i as heldOutGate, t as loopProvenanceSpans, p as paretoPolicy, j as paretoSignificanceGate, u as provenanceRecordPath, v as provenanceSpansPath, r as runEval, w as surfaceContentHash } from '../provenance-BEITkFII.js';
 import { E as EProcessState, a as PairedBootstrapResult } from '../statistics-xP-cWc5k.js';
 import { L as LlmClientOptions } from '../llm-client-Bj7g0rqu.js';
 import { AgentProfile } from '@tangle-network/agent-interface';
@@ -1446,6 +1446,18 @@ interface TraceAnalystProposerOptions {
 /** Wrap agent-eval's trace-analyst registry as a SurfaceProposer (prompt-tier). */
 declare function traceAnalystProposer(opts: TraceAnalystProposerOptions): SurfaceProposer;
 
+/** The shared, out-of-repo root for campaign/benchmark run bundles. Keeping run
+ *  outputs here means they never land in a repo working tree (no per-repo
+ *  gitignore, no clutter, no accidental commits). Layout:
+ *    ~/.tangle/traces/<repo>/runs/<runName>/
+ *  where <repo> disambiguates runs across repos in one place. */
+declare function tangleTracesRoot(): string;
+/** Resolve a campaign `runDir`. An absolute path is honored as-is (the caller
+ *  chose an explicit location). A bare name is placed under the shared home root
+ *  so bundles never pollute a repo working tree — the default the harness should
+ *  compute so callers pass a *name*, not a path. */
+declare function resolveRunDir(runDir: string, repo?: string): string;
+
 /**
  * Shared campaign-score reductions used by every optimizer preset
  * (`runOptimization`, `runSkillOpt`, `compareProposers`). ONE definition of
@@ -1525,4 +1537,4 @@ declare function gitWorktreeAdapter(opts: GitWorktreeAdapterOptions): WorktreeAd
  *  as a ref under the adapter's worktree dir. */
 declare function resolveWorktreePath(surface: CodeSurface, worktreeDir?: string): string;
 
-export { type AcceptedEdit, type AceProposerOptions, type AnalystArtifact, type AnalystScenario, type ApplySkillPatchResult, type BuildAnalystSurfaceDispatchOptions, type CampaignBreakdown, CampaignResult, CampaignRunPlan, CampaignStorage, CodeSurface, type CompareProposersOptions, type DimensionRegression, DispatchContext, type EvalFixture, type EvalFixtureFile, type EvalFixtureLoadOptions, type EvalFixtureRunPlan, type EvalFixtureScenario, type EvalFixtureValidationMode, type FailureModeRecallJudgeOptions, type FapoAttributionSignals, type FapoEntryConfig, type FapoFailureCluster, type FapoOptimizationLevel, type FapoProposerOptions, type FapoReviewInput, type FapoReviewIssue, type FapoReviewResult, type FapoScopeContract, FsLabeledScenarioStore, type FsLabeledScenarioStoreOptions, Gate, GenerationRecord, type GitWorktreeAdapterOptions, type HaloProposerOptions, type HeldoutSignificance, type HeldoutSignificanceOptions, type JsonPrimitive, type JsonValue, JudgeConfig, JudgeScore, LabelTrust, LabeledScenarioRecord, LabeledScenarioSampleArgs, LabeledScenarioSource, LabeledScenarioStore, LabeledScenarioStoreError, LabeledScenarioWrite, type LoadEvalFixtureScenariosOptions, type MemoryCurationProposerOptions, MutableSurface, type OptimizerEntryConfig, type PairedHoldout, type ParameterCandidate, type ParameterChange, type ParameterSweepProposerOptions, PlanCampaignRunOptions, type PlanEvalFixtureRunOptions, type PlaybackContext, type PlaybackDriver, type PlaybackStep, type PolicyEditProposerOptions, type ProfileDispatchFn, ProfileMatrixError, type ProfileSummary, ProposeContext, type ProposePatchesArgs, ProposedCandidate, type ProposerComparison, type ProposerEntry, type ProposerPairwise, type ProposerScore, type RejectedEdit, RunCampaignOptions, RunImprovementLoopOptions, type RunProfileMatrixOptions, type RunProfileMatrixResult, type RunSkillOptOptions, type RunSkillOptResult, Scenario, type ScenarioRollup, type ScoreboardRenderOptions, type ScoreboardRow, type ScoreboardSummary, type SequentialDecideFn, type SequentialDecideOptions, type SequentialDecision, type SequentialObservation, type SequentialPairedGate, type SequentialPairedGateOptions, type SkillOptEpochRecord, type SkillOptEvidence, type SkillOptProposer, type SkillOptProposerOptions, type SkillPatch, type SkillPatchOp, SkillPatchParseError, type SkillPatchRejection, SurfaceProposer, type TraceAnalystProposerOptions, type UserStory, type UserStoryVerdict, type Worktree, type WorktreeAdapter, WorktreeAdapterError, aceProposer, applySkillPatch, buildAnalystSurfaceDispatch, campaignBreakdown, campaignMeanComposite, compareProposers, detectScale, dimensionRegressions, discoverEvalFixtures, extractFapoAttributionSignals, failureModeRecallJudge, fapoEscalationEntry, fapoProposer, gepaParetoEntry, gepaReflectionEntry, gitWorktreeAdapter, haloProposer, heldoutSignificance, loadEvalFixture, loadEvalFixtureScenarios, makePlaybackDispatch, memoryCurationProposer, pairHoldout, parameterSweepProposer, parseSkillPatchResponse, patchEditCount, planEvalFixtureRun, policyEditProposer, renderScoreboardMarkdown, resolveWorktreePath, runProfileMatrix, runSkillOpt, scoreUserStory, scoreboardSummary, sequentialDecide, sequentialPairedGate, skillOptEntry, skillOptProposer, traceAnalystProposer, userStoryScoreboard };
+export { type AcceptedEdit, type AceProposerOptions, type AnalystArtifact, type AnalystScenario, type ApplySkillPatchResult, type BuildAnalystSurfaceDispatchOptions, type CampaignBreakdown, CampaignResult, CampaignRunPlan, CampaignStorage, CodeSurface, type CompareProposersOptions, type DimensionRegression, DispatchContext, type EvalFixture, type EvalFixtureFile, type EvalFixtureLoadOptions, type EvalFixtureRunPlan, type EvalFixtureScenario, type EvalFixtureValidationMode, type FailureModeRecallJudgeOptions, type FapoAttributionSignals, type FapoEntryConfig, type FapoFailureCluster, type FapoOptimizationLevel, type FapoProposerOptions, type FapoReviewInput, type FapoReviewIssue, type FapoReviewResult, type FapoScopeContract, FsLabeledScenarioStore, type FsLabeledScenarioStoreOptions, Gate, GenerationRecord, type GitWorktreeAdapterOptions, type HaloProposerOptions, type HeldoutSignificance, type HeldoutSignificanceOptions, type JsonPrimitive, type JsonValue, JudgeConfig, JudgeScore, LabelTrust, LabeledScenarioRecord, LabeledScenarioSampleArgs, LabeledScenarioSource, LabeledScenarioStore, LabeledScenarioStoreError, LabeledScenarioWrite, type LoadEvalFixtureScenariosOptions, type MemoryCurationProposerOptions, MutableSurface, type OptimizerEntryConfig, type PairedHoldout, type ParameterCandidate, type ParameterChange, type ParameterSweepProposerOptions, PlanCampaignRunOptions, type PlanEvalFixtureRunOptions, type PlaybackContext, type PlaybackDriver, type PlaybackStep, type PolicyEditProposerOptions, type ProfileDispatchFn, ProfileMatrixError, type ProfileSummary, ProposeContext, type ProposePatchesArgs, ProposedCandidate, type ProposerComparison, type ProposerEntry, type ProposerPairwise, type ProposerScore, type RejectedEdit, RunCampaignOptions, RunImprovementLoopOptions, type RunProfileMatrixOptions, type RunProfileMatrixResult, type RunSkillOptOptions, type RunSkillOptResult, Scenario, type ScenarioRollup, type ScoreboardRenderOptions, type ScoreboardRow, type ScoreboardSummary, type SequentialDecideFn, type SequentialDecideOptions, type SequentialDecision, type SequentialObservation, type SequentialPairedGate, type SequentialPairedGateOptions, type SkillOptEpochRecord, type SkillOptEvidence, type SkillOptProposer, type SkillOptProposerOptions, type SkillPatch, type SkillPatchOp, SkillPatchParseError, type SkillPatchRejection, SurfaceProposer, type TraceAnalystProposerOptions, type UserStory, type UserStoryVerdict, type Worktree, type WorktreeAdapter, WorktreeAdapterError, aceProposer, applySkillPatch, buildAnalystSurfaceDispatch, campaignBreakdown, campaignMeanComposite, compareProposers, detectScale, dimensionRegressions, discoverEvalFixtures, extractFapoAttributionSignals, failureModeRecallJudge, fapoEscalationEntry, fapoProposer, gepaParetoEntry, gepaReflectionEntry, gitWorktreeAdapter, haloProposer, heldoutSignificance, loadEvalFixture, loadEvalFixtureScenarios, makePlaybackDispatch, memoryCurationProposer, pairHoldout, parameterSweepProposer, parseSkillPatchResponse, patchEditCount, planEvalFixtureRun, policyEditProposer, renderScoreboardMarkdown, resolveRunDir, resolveWorktreePath, runProfileMatrix, runSkillOpt, scoreUserStory, scoreboardSummary, sequentialDecide, sequentialPairedGate, skillOptEntry, skillOptProposer, tangleTracesRoot, traceAnalystProposer, userStoryScoreboard };

package/dist/campaign/index.js

@@ -10,7 +10,7 @@ import {
   paretoPolicy,
   paretoSignificanceGate,
   runEval
-} from "../chunk-CHIFZIQD.js";
+} from "../chunk-QUCGGMYM.js";
 import {
   agentProfileHash,
   agentProfileId,
@@ -19,7 +19,7 @@ import {
   harnessAxisOf,
   llmJudge,
   verifyCompletion
-} from "../chunk-GUII3E73.js";
+} from "../chunk-CMJSTXUR.js";
 import {
   buildLoopProvenanceRecord,
   campaignBreakdown,
@@ -41,16 +41,18 @@ import {
   runOptimization,
   surfaceContentHash,
   surfaceHash
-} from "../chunk-NK77GPUH.js";
+} from "../chunk-52CCCXU3.js";
 import {
   assertRealBackend,
   contentHash,
   fsCampaignStorage,
   inMemoryCampaignStorage,
   planCampaignRun,
+  resolveRunDir,
   runCampaign,
-  summarizeBackendIntegrity
-} from "../chunk-XIOQHCHU.js";
+  summarizeBackendIntegrity,
+  tangleTracesRoot
+} from "../chunk-LSCBODPQ.js";
 import {
   estimateCost,
   isModelPriced
@@ -2756,6 +2758,7 @@ export {
   provenanceRecordPath,
   provenanceSpansPath,
   renderScoreboardMarkdown,
+  resolveRunDir,
   resolveWorktreePath,
   runCampaign,
   runEval,
@@ -2771,6 +2774,7 @@ export {
   skillOptProposer,
   surfaceContentHash,
   surfaceHash,
+  tangleTracesRoot,
   traceAnalystProposer,
   userStoryScoreboard
 };