engineering-behavior-observatory 0.2.1

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 (178) hide show
  1. package/LICENSE +202 -0
  2. package/README.md +151 -0
  3. package/contracts/codex-app-server-0.150.1/manifest.json +25 -0
  4. package/contracts/codex-app-server-0.150.1/schema/InitializeParams.json +84 -0
  5. package/contracts/codex-app-server-0.150.1/schema/JSONRPCMessage.json +137 -0
  6. package/contracts/codex-app-server-0.150.1/schema/v2/ThreadReadParams.json +17 -0
  7. package/contracts/codex-app-server-0.150.1/schema/v2/ThreadTokenUsageUpdatedNotification.json +82 -0
  8. package/contracts/codex-app-server-0.150.1/schema/v2/TurnInterruptParams.json +17 -0
  9. package/contracts/codex-app-server-0.150.1/types/AskForApproval.ts +5 -0
  10. package/contracts/codex-app-server-0.150.1/types/ClientNotification.ts +5 -0
  11. package/contracts/codex-app-server-0.150.1/types/SandboxMode.ts +5 -0
  12. package/contracts/codex-app-server-0.150.1/types/ThreadReadParams.ts +9 -0
  13. package/contracts/codex-app-server-0.150.1/types/TokenUsageBreakdown.ts +5 -0
  14. package/contracts/codex-app-server-0.150.1/types/TurnInterruptParams.ts +5 -0
  15. package/contracts/codex-app-server-0.153.4/manifest.json +27 -0
  16. package/contracts/codex-app-server-0.153.4/schema/JSONRPCMessage.json +137 -0
  17. package/contracts/codex-app-server-0.153.4/schema/v1/InitializeParams.json +84 -0
  18. package/contracts/codex-app-server-0.153.4/schema/v2/ThreadReadParams.json +17 -0
  19. package/contracts/codex-app-server-0.153.4/schema/v2/ThreadStartParams.json +515 -0
  20. package/contracts/codex-app-server-0.153.4/schema/v2/ThreadTokenUsageUpdatedNotification.json +82 -0
  21. package/contracts/codex-app-server-0.153.4/schema/v2/TurnInterruptParams.json +17 -0
  22. package/contracts/codex-app-server-0.153.4/schema/v2/TurnStartParams.json +921 -0
  23. package/contracts/codex-app-server-0.153.4/types/AskForApproval.ts +5 -0
  24. package/contracts/codex-app-server-0.153.4/types/ClientNotification.ts +5 -0
  25. package/contracts/codex-app-server-0.153.4/types/SandboxMode.ts +5 -0
  26. package/contracts/codex-app-server-0.153.4/types/ThreadReadParams.ts +12 -0
  27. package/contracts/codex-app-server-0.153.4/types/TokenUsageBreakdown.ts +5 -0
  28. package/contracts/codex-app-server-0.153.4/types/TurnInterruptParams.ts +5 -0
  29. package/contracts/openhands-agent-server-v1.44.1.json +65 -0
  30. package/contracts/openhands-agent-server-v1.46.0.json +46 -0
  31. package/dist/contracts/codex-app-server-0.153.4/types/AskForApproval.d.ts +9 -0
  32. package/dist/contracts/codex-app-server-0.153.4/types/AskForApproval.js +2 -0
  33. package/dist/contracts/codex-app-server-0.153.4/types/ClientNotification.d.ts +3 -0
  34. package/dist/contracts/codex-app-server-0.153.4/types/ClientNotification.js +2 -0
  35. package/dist/contracts/codex-app-server-0.153.4/types/SandboxMode.d.ts +1 -0
  36. package/dist/contracts/codex-app-server-0.153.4/types/SandboxMode.js +2 -0
  37. package/dist/contracts/codex-app-server-0.153.4/types/ThreadReadParams.d.ts +10 -0
  38. package/dist/contracts/codex-app-server-0.153.4/types/ThreadReadParams.js +2 -0
  39. package/dist/contracts/codex-app-server-0.153.4/types/TokenUsageBreakdown.d.ts +8 -0
  40. package/dist/contracts/codex-app-server-0.153.4/types/TokenUsageBreakdown.js +2 -0
  41. package/dist/contracts/codex-app-server-0.153.4/types/TurnInterruptParams.d.ts +4 -0
  42. package/dist/contracts/codex-app-server-0.153.4/types/TurnInterruptParams.js +2 -0
  43. package/dist/src/agent-sdk-normalizer.d.ts +22 -0
  44. package/dist/src/agent-sdk-normalizer.js +995 -0
  45. package/dist/src/agent-sdk-run.d.ts +31 -0
  46. package/dist/src/agent-sdk-run.js +242 -0
  47. package/dist/src/agent-sdk-runner.d.ts +96 -0
  48. package/dist/src/agent-sdk-runner.js +407 -0
  49. package/dist/src/agent-sdk.d.ts +294 -0
  50. package/dist/src/agent-sdk.js +644 -0
  51. package/dist/src/aggregation.d.ts +183 -0
  52. package/dist/src/aggregation.js +513 -0
  53. package/dist/src/artifacts.d.ts +36 -0
  54. package/dist/src/artifacts.js +1372 -0
  55. package/dist/src/atlas-grafana.d.ts +377 -0
  56. package/dist/src/atlas-grafana.js +47 -0
  57. package/dist/src/atlas-html.d.ts +2 -0
  58. package/dist/src/atlas-html.js +34 -0
  59. package/dist/src/atlas.d.ts +130 -0
  60. package/dist/src/atlas.js +235 -0
  61. package/dist/src/behavior-assertions.d.ts +95 -0
  62. package/dist/src/behavior-assertions.js +154 -0
  63. package/dist/src/cli.d.ts +5 -0
  64. package/dist/src/cli.js +957 -0
  65. package/dist/src/codex-judge.d.ts +44 -0
  66. package/dist/src/codex-judge.js +214 -0
  67. package/dist/src/codex-run.d.ts +114 -0
  68. package/dist/src/codex-run.js +519 -0
  69. package/dist/src/codex.d.ts +187 -0
  70. package/dist/src/codex.js +1028 -0
  71. package/dist/src/contracts.d.ts +155 -0
  72. package/dist/src/contracts.js +866 -0
  73. package/dist/src/corpus.d.ts +67 -0
  74. package/dist/src/corpus.js +540 -0
  75. package/dist/src/cursor-sdk-runner.d.ts +81 -0
  76. package/dist/src/cursor-sdk-runner.js +362 -0
  77. package/dist/src/cursor-sdk.d.ts +81 -0
  78. package/dist/src/cursor-sdk.js +1078 -0
  79. package/dist/src/deepseek-adapter.d.ts +152 -0
  80. package/dist/src/deepseek-adapter.js +777 -0
  81. package/dist/src/exports.d.ts +69 -0
  82. package/dist/src/exports.js +966 -0
  83. package/dist/src/human-calibration.d.ts +167 -0
  84. package/dist/src/human-calibration.js +618 -0
  85. package/dist/src/index.d.ts +71 -0
  86. package/dist/src/index.js +38 -0
  87. package/dist/src/lifecycle.d.ts +196 -0
  88. package/dist/src/lifecycle.js +1889 -0
  89. package/dist/src/normalization-integrity.d.ts +128 -0
  90. package/dist/src/normalization-integrity.js +429 -0
  91. package/dist/src/openhands-run.d.ts +36 -0
  92. package/dist/src/openhands-run.js +415 -0
  93. package/dist/src/openhands.d.ts +130 -0
  94. package/dist/src/openhands.js +859 -0
  95. package/dist/src/pi.d.ts +203 -0
  96. package/dist/src/pi.js +1345 -0
  97. package/dist/src/process-protocol.d.ts +245 -0
  98. package/dist/src/process-protocol.js +1261 -0
  99. package/dist/src/retained-evidence.d.ts +12 -0
  100. package/dist/src/retained-evidence.js +201 -0
  101. package/dist/src/run-bundles.d.ts +193 -0
  102. package/dist/src/run-bundles.js +993 -0
  103. package/dist/src/scheduler.d.ts +109 -0
  104. package/dist/src/scheduler.js +1080 -0
  105. package/dist/src/semantic-judge.d.ts +194 -0
  106. package/dist/src/semantic-judge.js +875 -0
  107. package/dist/src/structural-observations.d.ts +124 -0
  108. package/dist/src/structural-observations.js +671 -0
  109. package/dist/src/task-packets.d.ts +115 -0
  110. package/dist/src/task-packets.js +683 -0
  111. package/dist/src/uniform-events.d.ts +120 -0
  112. package/dist/src/uniform-events.js +158 -0
  113. package/dist/src/verifiers.d.ts +116 -0
  114. package/dist/src/verifiers.js +819 -0
  115. package/dist/src/workspaces.d.ts +56 -0
  116. package/dist/src/workspaces.js +1314 -0
  117. package/docs/README.md +60 -0
  118. package/docs/development/README.md +71 -0
  119. package/docs/development/documentation-sync.md +58 -0
  120. package/docs/development/extension-contracts.md +307 -0
  121. package/docs/evaluation/README.md +48 -0
  122. package/docs/evaluation/aggregation.md +112 -0
  123. package/docs/evaluation/behavior-assertions.md +45 -0
  124. package/docs/evaluation/human-calibration.md +170 -0
  125. package/docs/evaluation/normalization-integrity.md +80 -0
  126. package/docs/evaluation/semantic-judge.md +188 -0
  127. package/docs/evaluation/structural-observations.md +84 -0
  128. package/docs/evaluation/uniform-events.md +131 -0
  129. package/docs/guides/atlas.md +144 -0
  130. package/docs/guides/evidence-and-sharing.md +100 -0
  131. package/docs/guides/operator-guide.md +424 -0
  132. package/docs/guides/quickstart.md +152 -0
  133. package/docs/guides/telemetry.md +66 -0
  134. package/docs/harnesses/README.md +31 -0
  135. package/docs/harnesses/claude-agent-sdk.md +61 -0
  136. package/docs/harnesses/codex-harness.md +174 -0
  137. package/docs/harnesses/cursor-sdk.md +149 -0
  138. package/docs/harnesses/deepseek-harness.md +134 -0
  139. package/docs/harnesses/openhands-agent-server.md +153 -0
  140. package/docs/harnesses/pi-sdk.md +66 -0
  141. package/docs/reference/README.md +19 -0
  142. package/docs/reference/agent-sdk-operational-runner.md +118 -0
  143. package/docs/reference/cli.md +114 -0
  144. package/docs/reference/contracts.md +222 -0
  145. package/docs/reference/run-bundle-contract.md +354 -0
  146. package/docs/reference/run-lifecycle.md +54 -0
  147. package/examples/cursor-sdk/README.md +11 -0
  148. package/examples/cursor-sdk/capture-profile.json +1 -0
  149. package/examples/cursor-sdk/harness.json +1 -0
  150. package/examples/cursor-sdk/model.json +1 -0
  151. package/examples/cursor-sdk/native-limits.json +1 -0
  152. package/examples/cursor-sdk/native-tool-policy.json +1 -0
  153. package/ontology/behavior-categories.v1.json +46 -0
  154. package/package.json +66 -0
  155. package/release/0.1.0/KNOWN_LIMITATIONS.md +35 -0
  156. package/release/0.1.0/README.md +46 -0
  157. package/release/0.1.0/reproducibility.json +61 -0
  158. package/release/0.2.0/KNOWN_LIMITATIONS.md +40 -0
  159. package/release/0.2.0/README.md +42 -0
  160. package/release/0.2.0/reproducibility.json +73 -0
  161. package/release/0.2.1/KNOWN_LIMITATIONS.md +18 -0
  162. package/release/0.2.1/README.md +63 -0
  163. package/release/0.2.1/reproducibility.json +73 -0
  164. package/release/README.md +13 -0
  165. package/schemas/aggregation.v1.json +293 -0
  166. package/schemas/behavior-assertions.v1.json +220 -0
  167. package/schemas/deepseek-runtime-composition.v1.schema.json +122 -0
  168. package/schemas/experiment.v1.schema.json +188 -0
  169. package/schemas/human-calibration.v1.json +336 -0
  170. package/schemas/normalization-integrity.v1.json +302 -0
  171. package/schemas/run-bundles/v1.json +1040 -0
  172. package/schemas/run-queue.v1.schema.json +206 -0
  173. package/schemas/semantic-judge.v1.json +380 -0
  174. package/schemas/structural-observations.v1.json +131 -0
  175. package/schemas/task-packet-freeze.v1.schema.json +106 -0
  176. package/schemas/task-packet.v1.schema.json +234 -0
  177. package/schemas/uniform-events/v1.json +289 -0
  178. package/scripts/atlas-grafana.sh +18 -0
@@ -0,0 +1,183 @@
1
+ import { type BehaviorAssertion } from "./behavior-assertions.js";
2
+ import type { CorpusIndexEntry } from "./corpus.js";
3
+ import { effectiveReviewOutcome, type ReviewHistory, type ReviewSample } from "./human-calibration.js";
4
+ import { type ComparisonReport, type ComparisonRequest } from "./normalization-integrity.js";
5
+ import { type StructuralObservationSet } from "./structural-observations.js";
6
+ export type AggregationDimension = "task" | "model" | "harness" | "trial" | "capture-qualification";
7
+ export type AttemptSelectionPolicy = "all-attempts" | "latest-attempt-per-run";
8
+ export type AggregationRequest = {
9
+ schemaVersion: "ebo.aggregation-request/v1";
10
+ sources: {
11
+ corpusRoot: string;
12
+ corpusIndex: string;
13
+ observationSets: ReadonlyArray<{
14
+ bundleRoot: string;
15
+ path: string;
16
+ }>;
17
+ assertions: ReadonlyArray<{
18
+ bundleRoot: string;
19
+ path: string;
20
+ }>;
21
+ calibrations: ReadonlyArray<{
22
+ selection: string;
23
+ history: string;
24
+ }>;
25
+ };
26
+ groupBy: readonly AggregationDimension[];
27
+ selectedAttemptPolicy: AttemptSelectionPolicy;
28
+ recurrence: {
29
+ minimumOccurrences: number;
30
+ };
31
+ comparisons: ReadonlyArray<{
32
+ id: string;
33
+ measure: string;
34
+ left: Partial<Record<AggregationDimension, string>>;
35
+ right: Partial<Record<AggregationDimension, string>>;
36
+ matchBy: readonly AggregationDimension[];
37
+ eligibilityGates: ReadonlyArray<{
38
+ request: string;
39
+ report: string;
40
+ }>;
41
+ }>;
42
+ };
43
+ export type AggregationInput = {
44
+ lineage?: {
45
+ requestDigest?: `sha256:${string}`;
46
+ corpusIndexDigest?: `sha256:${string}`;
47
+ };
48
+ corpusEntries: readonly CorpusIndexEntry[];
49
+ observationSets: ReadonlyArray<{
50
+ bundleRoot: string;
51
+ document: StructuralObservationSet;
52
+ }>;
53
+ assertions: ReadonlyArray<{
54
+ bundleRoot: string;
55
+ document: BehaviorAssertion;
56
+ }>;
57
+ calibrations: ReadonlyArray<{
58
+ selection: ReviewSample;
59
+ history: ReviewHistory;
60
+ }>;
61
+ comparisons: ReadonlyArray<{
62
+ id: string;
63
+ measure: string;
64
+ left: Partial<Record<AggregationDimension, string>>;
65
+ right: Partial<Record<AggregationDimension, string>>;
66
+ matchBy: readonly AggregationDimension[];
67
+ eligibility: ReadonlyArray<{
68
+ request: ComparisonRequest;
69
+ report: ComparisonReport;
70
+ }>;
71
+ }>;
72
+ };
73
+ export type AggregateMeasurement = {
74
+ status: "available" | "unavailable";
75
+ numerator: {
76
+ value: number;
77
+ unit: string;
78
+ };
79
+ denominator: {
80
+ value: number;
81
+ unit: string;
82
+ };
83
+ rate?: number;
84
+ exclusions: ReadonlyArray<{
85
+ reason: string;
86
+ count: number;
87
+ unit: string;
88
+ }>;
89
+ reason?: string;
90
+ };
91
+ export type AggregateMetric = {
92
+ id: string;
93
+ population: "run" | "attempt" | "operation" | "assertion" | "reviewed-assertion";
94
+ claimStatus: "descriptive" | "unavailable";
95
+ measurement: AggregateMeasurement;
96
+ };
97
+ export type BehaviorAggregate = {
98
+ behavior: BehaviorAssertion["behavior"];
99
+ rubric: BehaviorAssertion["rubric"];
100
+ evaluator: BehaviorAssertion["evaluator"];
101
+ population: "confirmed-attempt-dimension";
102
+ assessments: ReadonlyArray<{
103
+ assessment: "constructive" | "adverse" | "mixed" | "context-dependent";
104
+ measurement: AggregateMeasurement;
105
+ }>;
106
+ assertions: ReadonlyArray<{
107
+ runId: string;
108
+ attemptId: string;
109
+ id: string;
110
+ digest: `sha256:${string}`;
111
+ reviewOutcome: ReturnType<typeof effectiveReviewOutcome>;
112
+ disputed: boolean;
113
+ included: boolean;
114
+ }>;
115
+ };
116
+ export type AggregationReport = {
117
+ schemaVersion: "ebo.aggregation-report/v1";
118
+ policy: {
119
+ groupBy: readonly AggregationDimension[];
120
+ selectedAttemptPolicy: AttemptSelectionPolicy;
121
+ recurrence: {
122
+ minimumOccurrences: number;
123
+ establishesCausality: false;
124
+ establishesStatisticalSignificance: false;
125
+ };
126
+ };
127
+ sourcePopulation: {
128
+ uniqueRuns: number;
129
+ uniqueAttempts: number;
130
+ selectedAttempts: number;
131
+ duplicateInputsIgnored: number;
132
+ };
133
+ sourceLineage: {
134
+ requestDigest: `sha256:${string}`;
135
+ corpusIndexDigest: `sha256:${string}`;
136
+ manifests: ReadonlyArray<{
137
+ runId: string;
138
+ attemptId: string;
139
+ digest: `sha256:${string}`;
140
+ }>;
141
+ observationSets: ReadonlyArray<{
142
+ runId: string;
143
+ attemptId: string;
144
+ digest: `sha256:${string}`;
145
+ }>;
146
+ assertions: ReadonlyArray<{
147
+ runId: string;
148
+ attemptId: string;
149
+ id: string;
150
+ digest: `sha256:${string}`;
151
+ }>;
152
+ calibrations: ReadonlyArray<{
153
+ selectionDigest: `sha256:${string}`;
154
+ historyDigest: `sha256:${string}`;
155
+ }>;
156
+ comparisonGates: ReadonlyArray<{
157
+ requestDigest: `sha256:${string}`;
158
+ reportDigest: `sha256:${string}`;
159
+ }>;
160
+ };
161
+ groups: ReadonlyArray<{
162
+ dimensions: Partial<Record<AggregationDimension, string>>;
163
+ metrics: readonly AggregateMetric[];
164
+ behaviors?: readonly BehaviorAggregate[];
165
+ variations: ReadonlyArray<{
166
+ measure: "terminal-state" | "verifier-status";
167
+ distribution: readonly AggregateMetric[];
168
+ claimStatus: "no-variation" | "case-study" | "recurring-description" | "unavailable";
169
+ }>;
170
+ }>;
171
+ comparisons: ReadonlyArray<{
172
+ id: string;
173
+ measure: string;
174
+ eligibility: readonly ComparisonReport[];
175
+ matchedDifference: AggregateMeasurement;
176
+ differingMatchedUnits: number;
177
+ claimStatus: "no-difference" | "case-study" | "recurring-description" | "unavailable";
178
+ limitations: readonly string[];
179
+ }>;
180
+ limitations: readonly string[];
181
+ };
182
+ export declare function aggregateEvaluation(input: AggregationInput, policy: Pick<AggregationRequest, "groupBy" | "selectedAttemptPolicy" | "recurrence">): Promise<AggregationReport>;
183
+ export declare function selectAggregationAttempts(entries: readonly CorpusIndexEntry[], policy: AttemptSelectionPolicy): CorpusIndexEntry[];