@tangle-network/agent-eval 0.125.0 → 0.126.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (82) hide show
  1. package/CHANGELOG.md +62 -35
  2. package/README.md +270 -189
  3. package/dist/analyst/index.d.ts +15 -145
  4. package/dist/analyst/index.js +33 -47
  5. package/dist/analyst/index.js.map +1 -1
  6. package/dist/benchmarks/index.d.ts +45 -162
  7. package/dist/benchmarks/index.js +8 -9
  8. package/dist/campaign/index.d.ts +3674 -5393
  9. package/dist/campaign/index.js +21 -95
  10. package/dist/{chunk-R226UZOI.js → chunk-474LBSOX.js} +2 -2
  11. package/dist/{chunk-HM6V7F3M.js → chunk-FO7HEH76.js} +3 -3
  12. package/dist/chunk-IILEIWGW.js +635 -0
  13. package/dist/chunk-IILEIWGW.js.map +1 -0
  14. package/dist/{chunk-EQUK3RFS.js → chunk-J5SQWP6Y.js} +8 -5
  15. package/dist/chunk-J5SQWP6Y.js.map +1 -0
  16. package/dist/{chunk-W5B3ZGP3.js → chunk-KE2VWPZX.js} +8 -6
  17. package/dist/{chunk-W5B3ZGP3.js.map → chunk-KE2VWPZX.js.map} +1 -1
  18. package/dist/{chunk-DT7OXY3C.js → chunk-LUNF2SEL.js} +538 -851
  19. package/dist/chunk-LUNF2SEL.js.map +1 -0
  20. package/dist/chunk-NGUYT5CI.js +4637 -0
  21. package/dist/chunk-NGUYT5CI.js.map +1 -0
  22. package/dist/{chunk-QFQZ3U3X.js → chunk-OCFJACJU.js} +2 -2
  23. package/dist/{chunk-GID26AN4.js → chunk-P22LJ3Y2.js} +4 -6
  24. package/dist/{chunk-GID26AN4.js.map → chunk-P22LJ3Y2.js.map} +1 -1
  25. package/dist/{chunk-SJT4OBVL.js → chunk-SDPM6554.js} +3 -3
  26. package/dist/{chunk-D5JZ7UDZ.js → chunk-UCLVDLCH.js} +136 -50
  27. package/dist/chunk-UCLVDLCH.js.map +1 -0
  28. package/dist/chunk-VMUENW6F.js +7274 -0
  29. package/dist/chunk-VMUENW6F.js.map +1 -0
  30. package/dist/{chunk-JKDNAOF5.js → chunk-W4L6C2XT.js} +2 -2
  31. package/dist/chunk-WGXIEX7P.js +116 -0
  32. package/dist/chunk-WGXIEX7P.js.map +1 -0
  33. package/dist/{chunk-GRCDRKII.js → chunk-WS3NZZQQ.js} +58 -20
  34. package/dist/chunk-WS3NZZQQ.js.map +1 -0
  35. package/dist/cli.js +3 -3
  36. package/dist/contract/index.d.ts +3220 -3094
  37. package/dist/contract/index.js +173 -42
  38. package/dist/contract/index.js.map +1 -1
  39. package/dist/control.js +2 -3
  40. package/dist/fuzz.d.ts +14 -1
  41. package/dist/fuzz.js +1 -1
  42. package/dist/hosted/index.d.ts +8 -1
  43. package/dist/index.d.ts +71 -687
  44. package/dist/index.js +178 -497
  45. package/dist/index.js.map +1 -1
  46. package/dist/openapi.json +1 -1
  47. package/dist/rl.d.ts +5 -100
  48. package/dist/rl.js +4 -5
  49. package/dist/rl.js.map +1 -1
  50. package/dist/{run-campaign-I3JXKVAK.js → run-campaign-LVFKZCEU.js} +3 -3
  51. package/dist/traces.js +2 -3
  52. package/dist/wire/index.d.ts +14 -1
  53. package/dist/wire/index.js +3 -3
  54. package/docs/campaign-proposers.md +363 -168
  55. package/docs/design/loop-taxonomy.md +142 -190
  56. package/docs/design.md +1 -1
  57. package/docs/distributed-driver.md +8 -11
  58. package/docs/feature-guide.md +20 -19
  59. package/docs/knowledge-readiness.md +2 -5
  60. package/docs/multi-shot-optimization.md +35 -27
  61. package/docs/rollout.md +5 -5
  62. package/package.json +4 -4
  63. package/dist/chunk-A62YMFWA.js +0 -9269
  64. package/dist/chunk-A62YMFWA.js.map +0 -1
  65. package/dist/chunk-A6GT67HT.js +0 -550
  66. package/dist/chunk-A6GT67HT.js.map +0 -1
  67. package/dist/chunk-D5JZ7UDZ.js.map +0 -1
  68. package/dist/chunk-DT7OXY3C.js.map +0 -1
  69. package/dist/chunk-EQUK3RFS.js.map +0 -1
  70. package/dist/chunk-GC4ATIKK.js +0 -317
  71. package/dist/chunk-GC4ATIKK.js.map +0 -1
  72. package/dist/chunk-GRCDRKII.js.map +0 -1
  73. package/dist/chunk-LOW3U7JZ.js +0 -328
  74. package/dist/chunk-LOW3U7JZ.js.map +0 -1
  75. package/dist/chunk-PMITBABE.js +0 -3841
  76. package/dist/chunk-PMITBABE.js.map +0 -1
  77. /package/dist/{chunk-R226UZOI.js.map → chunk-474LBSOX.js.map} +0 -0
  78. /package/dist/{chunk-HM6V7F3M.js.map → chunk-FO7HEH76.js.map} +0 -0
  79. /package/dist/{chunk-QFQZ3U3X.js.map → chunk-OCFJACJU.js.map} +0 -0
  80. /package/dist/{chunk-SJT4OBVL.js.map → chunk-SDPM6554.js.map} +0 -0
  81. /package/dist/{chunk-JKDNAOF5.js.map → chunk-W4L6C2XT.js.map} +0 -0
  82. /package/dist/{run-campaign-I3JXKVAK.js.map → run-campaign-LVFKZCEU.js.map} +0 -0
@@ -1,79 +1,42 @@
1
1
  import {
2
- DEFAULT_POLICY_EDIT_HISTORY_LIMITS,
3
2
  FileSearchLedger,
4
3
  FsLabeledScenarioStore,
5
4
  LabeledScenarioStoreError,
6
- Lineage,
7
- LineageStoreConflictError,
8
5
  ProfileMatrixError,
9
6
  SEARCH_LEDGER_SCHEMA,
10
- SkillPatchParseError,
11
7
  WorktreeAdapterError,
12
- aceProposer,
13
- acquireSingleRunLock,
14
8
  analyzeCrossSurfaceInteractions,
15
- analyzeOtlpTraceFile,
16
- applySkillPatch,
17
- assertPolicyEditAuthorContextBudget,
18
9
  buildAnalystSurfaceDispatch,
19
- callbackGovernor,
20
- campaignLineageStore,
21
10
  classifyUngroundedLiterals,
22
- compareOptimizationMethods,
23
- compositeProposer,
24
- costFromLedgerSummary,
25
11
  discoverEvalFixtures,
26
- extractFapoAttributionSignals,
27
12
  failureModeRecallJudge,
28
- fapoEscalationMethod,
29
- fapoProposer,
30
- fsLineageStore,
31
- gepaOptimizationMethod,
32
- gepaParetoMethod,
33
- gepaReflectionMethod,
34
13
  gitWorktreeAdapter,
35
- haloProposer,
36
- heuristicGovernor,
37
14
  isTransientTransportFailure,
38
- lineageNodeId,
39
- llmPolicyEditProposer,
40
15
  loadEvalFixture,
41
16
  loadEvalFixtureScenarios,
42
17
  makePlaybackDispatch,
43
- memLineageStore,
44
- memoryCurationProposer,
45
18
  neutralizationGate,
46
19
  neutralizeText,
47
20
  openSearchLedger,
48
- parameterSweepProposer,
49
- parseSkillPatchResponse,
50
- patchEditCount,
51
21
  planEvalFixtureRun,
52
- policyEditProposer,
53
- projectPolicyEditHistory,
54
22
  renderScoreboardMarkdown,
55
23
  resolveWorktreePath,
56
24
  rolloutArgumentDiff,
57
- runLineage,
58
- runLineageLoop,
59
25
  runProfileMatrix,
60
- runSkillOpt,
61
26
  scoreDiscrimination,
62
27
  scoreUserStory,
63
28
  scoreboardSummary,
64
29
  selectDiscriminative,
65
- selectPolicyEditAuthorRows,
66
30
  sequentialDecide,
67
31
  sequentialPairedGate,
68
- skillOptMethod,
69
- skillOptProposer,
70
- traceAnalystProposer,
71
32
  userStoryScoreboard,
72
33
  validateSearchLedgerEvent,
73
34
  verifyCodeSurface
74
- } from "../chunk-A62YMFWA.js";
35
+ } from "../chunk-NGUYT5CI.js";
75
36
  import {
37
+ acquireSingleRunLock,
76
38
  assertCodeSurfaceIdentity,
39
+ assertComponentSurface,
77
40
  buildEvidenceVector,
78
41
  buildLoopProvenanceRecord,
79
42
  campaignBreakdown,
@@ -81,16 +44,17 @@ import {
81
44
  campaignMeasurementDigest,
82
45
  canonicalDigest,
83
46
  codeSurfaceIdentityMaterial,
47
+ compareOptimizationMethods,
48
+ componentSurfaceIdentityMaterial,
84
49
  composeGate,
85
- countSentenceEdits,
50
+ costFromLedgerSummary,
86
51
  createReferenceEquivalenceJudge,
87
52
  defaultProductionGate,
88
53
  detectScale,
89
54
  dimensionRegressions,
90
55
  emitLoopProvenance,
91
- evolutionaryProposer,
92
- extractH2Sections,
93
- gepaProposer,
56
+ externalTextOptimizationMethod,
57
+ gepaOptimizationMethod,
94
58
  heldOutGate,
95
59
  heldoutSignificance,
96
60
  isProposedCandidate,
@@ -99,6 +63,7 @@ import {
99
63
  loopProvenanceArgsFromResult,
100
64
  loopProvenanceSpans,
101
65
  openAutoPr,
66
+ optimizationTokenUsageFromSummary,
102
67
  pairHoldout,
103
68
  paretoPolicy,
104
69
  paretoSignificanceGate,
@@ -109,10 +74,11 @@ import {
109
74
  runEval,
110
75
  runImprovementLoop,
111
76
  runOptimization,
77
+ skillOptOptimizationMethod,
112
78
  surfaceContentHash,
113
79
  surfaceHash,
114
80
  verifyLoopProvenanceRecord
115
- } from "../chunk-PMITBABE.js";
81
+ } from "../chunk-VMUENW6F.js";
116
82
  import {
117
83
  SearchLedgerConflictError,
118
84
  SearchLedgerError,
@@ -129,57 +95,43 @@ import {
129
95
  resolveRunDir,
130
96
  runCampaign,
131
97
  tangleTracesRoot
132
- } from "../chunk-D5JZ7UDZ.js";
133
- import {
134
- POLICY_EDIT_CANDIDATE_RECORD_SCHEMA,
135
- validatePolicyEditCandidateRecord
136
- } from "../chunk-DT7OXY3C.js";
98
+ } from "../chunk-UCLVDLCH.js";
99
+ import "../chunk-WGXIEX7P.js";
137
100
  import "../chunk-ARU2PZFM.js";
138
- import "../chunk-EQUK3RFS.js";
101
+ import "../chunk-J5SQWP6Y.js";
139
102
  import "../chunk-PJQFMIOX.js";
140
- import "../chunk-GRCDRKII.js";
103
+ import "../chunk-WS3NZZQQ.js";
141
104
  import "../chunk-VI2UW6B6.js";
142
105
  import "../chunk-NUKSVU3W.js";
143
106
  import "../chunk-GGE4NNQT.js";
144
107
  import "../chunk-IR3KBHOY.js";
145
108
  import "../chunk-PC4UYEBM.js";
146
- import "../chunk-LOW3U7JZ.js";
109
+ import "../chunk-IILEIWGW.js";
147
110
  import "../chunk-MA6HLL3S.js";
148
- import "../chunk-GC4ATIKK.js";
149
111
  import "../chunk-VSMTAMNK.js";
150
112
  import "../chunk-ONWEPEDO.js";
151
113
  import "../chunk-K4DBDHLK.js";
152
114
  import "../chunk-PZ5AY32C.js";
153
115
  export {
154
- DEFAULT_POLICY_EDIT_HISTORY_LIMITS,
155
116
  FileSearchLedger,
156
117
  FsLabeledScenarioStore,
157
118
  LabeledScenarioStoreError,
158
- Lineage,
159
- LineageStoreConflictError,
160
- POLICY_EDIT_CANDIDATE_RECORD_SCHEMA,
161
119
  ProfileMatrixError,
162
120
  SEARCH_LEDGER_SCHEMA,
163
121
  SearchLedgerConflictError,
164
122
  SearchLedgerError,
165
123
  SearchLedgerIntegrityError,
166
- SkillPatchParseError,
167
124
  WorktreeAdapterError,
168
- aceProposer,
169
125
  acquireSingleRunLock,
170
126
  analyzeCrossSurfaceInteractions,
171
- analyzeOtlpTraceFile,
172
- applySkillPatch,
173
127
  assertCampaignDesign,
174
128
  assertCampaignSplitIdentity,
175
129
  assertCodeSurfaceIdentity,
176
- assertPolicyEditAuthorContextBudget,
130
+ assertComponentSurface,
177
131
  buildAnalystSurfaceDispatch,
178
132
  buildEvidenceVector,
179
133
  buildLoopProvenanceRecord,
180
- callbackGovernor,
181
134
  campaignBreakdown,
182
- campaignLineageStore,
183
135
  campaignMeanComposite,
184
136
  campaignMeasurementDigest,
185
137
  campaignScenarioIdentity,
@@ -189,10 +141,9 @@ export {
189
141
  classifyUngroundedLiterals,
190
142
  codeSurfaceIdentityMaterial,
191
143
  compareOptimizationMethods,
144
+ componentSurfaceIdentityMaterial,
192
145
  composeGate,
193
- compositeProposer,
194
146
  costFromLedgerSummary,
195
- countSentenceEdits,
196
147
  createReferenceEquivalenceJudge,
197
148
  createRunCostLedger,
198
149
  defaultProductionGate,
@@ -200,52 +151,34 @@ export {
200
151
  dimensionRegressions,
201
152
  discoverEvalFixtures,
202
153
  emitLoopProvenance,
203
- evolutionaryProposer,
204
- extractFapoAttributionSignals,
205
- extractH2Sections,
154
+ externalTextOptimizationMethod,
206
155
  failureModeRecallJudge,
207
- fapoEscalationMethod,
208
- fapoProposer,
209
156
  fsCampaignStorage,
210
- fsLineageStore,
211
157
  gepaOptimizationMethod,
212
- gepaParetoMethod,
213
- gepaProposer,
214
- gepaReflectionMethod,
215
158
  gitWorktreeAdapter,
216
- haloProposer,
217
159
  heldOutGate,
218
160
  heldoutSignificance,
219
- heuristicGovernor,
220
161
  inMemoryCampaignStorage,
221
162
  isProposedCandidate,
222
163
  isTransientTransportFailure,
223
164
  labelTrustRank,
224
- lineageNodeId,
225
165
  llmJudge,
226
- llmPolicyEditProposer,
227
166
  loadEvalFixture,
228
167
  loadEvalFixtureScenarios,
229
168
  loopProvenanceArgsFromResult,
230
169
  loopProvenanceSpans,
231
170
  makePlaybackDispatch,
232
- memLineageStore,
233
- memoryCurationProposer,
234
171
  neutralizationGate,
235
172
  neutralizeText,
236
173
  openAutoPr,
237
174
  openSearchLedger,
175
+ optimizationTokenUsageFromSummary,
238
176
  pairHoldout,
239
- parameterSweepProposer,
240
177
  paretoPolicy,
241
178
  paretoSignificanceGate,
242
- parseSkillPatchResponse,
243
- patchEditCount,
244
179
  planCampaignRun,
245
180
  planEvalFixtureRun,
246
- policyEditProposer,
247
181
  powerPreflight,
248
- projectPolicyEditHistory,
249
182
  provenanceRecordPath,
250
183
  provenanceSpansPath,
251
184
  renderScoreboardMarkdown,
@@ -256,26 +189,19 @@ export {
256
189
  runCampaign,
257
190
  runEval,
258
191
  runImprovementLoop,
259
- runLineage,
260
- runLineageLoop,
261
192
  runOptimization,
262
193
  runProfileMatrix,
263
- runSkillOpt,
264
194
  scoreDiscrimination,
265
195
  scoreUserStory,
266
196
  scoreboardSummary,
267
197
  selectDiscriminative,
268
- selectPolicyEditAuthorRows,
269
198
  sequentialDecide,
270
199
  sequentialPairedGate,
271
- skillOptMethod,
272
- skillOptProposer,
200
+ skillOptOptimizationMethod,
273
201
  surfaceContentHash,
274
202
  surfaceHash,
275
203
  tangleTracesRoot,
276
- traceAnalystProposer,
277
204
  userStoryScoreboard,
278
- validatePolicyEditCandidateRecord,
279
205
  validateSearchLedgerEvent,
280
206
  verifyCodeSurface,
281
207
  verifyLoopProvenanceRecord
@@ -3,7 +3,7 @@ import {
3
3
  } from "./chunk-VQMK5FMP.js";
4
4
  import {
5
5
  validateRunRecord
6
- } from "./chunk-LOW3U7JZ.js";
6
+ } from "./chunk-IILEIWGW.js";
7
7
 
8
8
  // src/action-policy.ts
9
9
  function evaluateActionPolicy(action, policy = {}, options = {}) {
@@ -1523,4 +1523,4 @@ export {
1523
1523
  runProposeReviewAsControlLoop,
1524
1524
  controlFailureClassFromVerification
1525
1525
  };
1526
- //# sourceMappingURL=chunk-R226UZOI.js.map
1526
+ //# sourceMappingURL=chunk-474LBSOX.js.map
@@ -5,10 +5,10 @@ import {
5
5
  costReceiptFromLlm,
6
6
  costReceiptFromLlmError,
7
7
  maximumChargeForLlmRequest
8
- } from "./chunk-EQUK3RFS.js";
8
+ } from "./chunk-J5SQWP6Y.js";
9
9
  import {
10
10
  CostLedger
11
- } from "./chunk-GRCDRKII.js";
11
+ } from "./chunk-WS3NZZQQ.js";
12
12
 
13
13
  // src/wire/schemas.ts
14
14
  import { extendZodWithOpenApi } from "@asteasolutions/zod-to-openapi";
@@ -1053,4 +1053,4 @@ export {
1053
1053
  startServer,
1054
1054
  startServerAsync
1055
1055
  };
1056
- //# sourceMappingURL=chunk-HM6V7F3M.js.map
1056
+ //# sourceMappingURL=chunk-FO7HEH76.js.map