@atlanai/sdk 0.1.0 → 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.
- package/README.md +94 -26
- package/dist/client.js +3 -1
- package/dist/evals.d.ts +46 -0
- package/dist/evals.js +230 -0
- package/dist/index.d.ts +3 -0
- package/dist/index.js +8 -1
- package/dist/manifest.json +299 -3
- package/dist/raw/eval/apis/DatasetsApi.d.ts +304 -0
- package/dist/raw/eval/apis/DatasetsApi.js +706 -0
- package/dist/raw/eval/apis/ExperimentsApi.d.ts +371 -0
- package/dist/raw/eval/apis/ExperimentsApi.js +869 -0
- package/dist/raw/eval/apis/ScorersApi.d.ts +129 -0
- package/dist/raw/eval/apis/ScorersApi.js +308 -0
- package/dist/raw/eval/apis/index.d.ts +3 -0
- package/dist/raw/eval/apis/index.js +21 -0
- package/dist/raw/eval/index.d.ts +3 -0
- package/dist/raw/eval/index.js +21 -0
- package/dist/raw/eval/models/AccessView.d.ts +57 -0
- package/dist/raw/eval/models/AccessView.js +64 -0
- package/dist/raw/eval/models/ArtifactConflictProblem.d.ts +69 -0
- package/dist/raw/eval/models/ArtifactConflictProblem.js +66 -0
- package/dist/raw/eval/models/CostView.d.ts +50 -0
- package/dist/raw/eval/models/CostView.js +60 -0
- package/dist/raw/eval/models/CursorPage.d.ts +38 -0
- package/dist/raw/eval/models/CursorPage.js +52 -0
- package/dist/raw/eval/models/EvalArtifactEnvelope.d.ts +144 -0
- package/dist/raw/eval/models/EvalArtifactEnvelope.js +84 -0
- package/dist/raw/eval/models/EvalArtifactSearchRequest.d.ts +71 -0
- package/dist/raw/eval/models/EvalArtifactSearchRequest.js +60 -0
- package/dist/raw/eval/models/EvalBulkDatasetRecordResponse.d.ts +33 -0
- package/dist/raw/eval/models/EvalBulkDatasetRecordResponse.js +51 -0
- package/dist/raw/eval/models/EvalBulkDatasetRecordResult.d.ts +52 -0
- package/dist/raw/eval/models/EvalBulkDatasetRecordResult.js +60 -0
- package/dist/raw/eval/models/EvalBulkExperimentResultResponse.d.ts +33 -0
- package/dist/raw/eval/models/EvalBulkExperimentResultResponse.js +51 -0
- package/dist/raw/eval/models/EvalBulkExperimentResultSlot.d.ts +53 -0
- package/dist/raw/eval/models/EvalBulkExperimentResultSlot.js +60 -0
- package/dist/raw/eval/models/EvalCreateDataset400Response.d.ts +65 -0
- package/dist/raw/eval/models/EvalCreateDataset400Response.js +75 -0
- package/dist/raw/eval/models/EvalCreateDataset409Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalCreateDataset409Response.js +75 -0
- package/dist/raw/eval/models/EvalCreateDataset413Response.d.ts +64 -0
- package/dist/raw/eval/models/EvalCreateDataset413Response.js +74 -0
- package/dist/raw/eval/models/EvalCreateDataset415Response.d.ts +64 -0
- package/dist/raw/eval/models/EvalCreateDataset415Response.js +74 -0
- package/dist/raw/eval/models/EvalCreateDataset422Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalCreateDataset422Response.js +75 -0
- package/dist/raw/eval/models/EvalDatasetArtifactView.d.ts +277 -0
- package/dist/raw/eval/models/EvalDatasetArtifactView.js +146 -0
- package/dist/raw/eval/models/EvalDatasetCreate.d.ts +150 -0
- package/dist/raw/eval/models/EvalDatasetCreate.js +86 -0
- package/dist/raw/eval/models/EvalDatasetMeta.d.ts +46 -0
- package/dist/raw/eval/models/EvalDatasetMeta.js +48 -0
- package/dist/raw/eval/models/EvalDatasetPatch.d.ts +52 -0
- package/dist/raw/eval/models/EvalDatasetPatch.js +54 -0
- package/dist/raw/eval/models/EvalDatasetRecordArtifactView.d.ts +330 -0
- package/dist/raw/eval/models/EvalDatasetRecordArtifactView.js +161 -0
- package/dist/raw/eval/models/EvalDatasetRecordBulkCreate.d.ts +33 -0
- package/dist/raw/eval/models/EvalDatasetRecordBulkCreate.js +51 -0
- package/dist/raw/eval/models/EvalDatasetRecordCreate.d.ts +195 -0
- package/dist/raw/eval/models/EvalDatasetRecordCreate.js +101 -0
- package/dist/raw/eval/models/EvalDatasetRecordMeta.d.ts +96 -0
- package/dist/raw/eval/models/EvalDatasetRecordMeta.js +67 -0
- package/dist/raw/eval/models/EvalDatasetRecordPatch.d.ts +81 -0
- package/dist/raw/eval/models/EvalDatasetRecordPatch.js +62 -0
- package/dist/raw/eval/models/EvalExperimentArtifactView.d.ts +320 -0
- package/dist/raw/eval/models/EvalExperimentArtifactView.js +160 -0
- package/dist/raw/eval/models/EvalExperimentCreate.d.ts +193 -0
- package/dist/raw/eval/models/EvalExperimentCreate.js +100 -0
- package/dist/raw/eval/models/EvalExperimentMeta.d.ts +91 -0
- package/dist/raw/eval/models/EvalExperimentMeta.js +62 -0
- package/dist/raw/eval/models/EvalExperimentPatch.d.ts +63 -0
- package/dist/raw/eval/models/EvalExperimentPatch.js +57 -0
- package/dist/raw/eval/models/EvalExperimentResultArtifactView.d.ts +330 -0
- package/dist/raw/eval/models/EvalExperimentResultArtifactView.js +162 -0
- package/dist/raw/eval/models/EvalExperimentResultBulkCreate.d.ts +33 -0
- package/dist/raw/eval/models/EvalExperimentResultBulkCreate.js +51 -0
- package/dist/raw/eval/models/EvalExperimentResultCreate.d.ts +197 -0
- package/dist/raw/eval/models/EvalExperimentResultCreate.js +100 -0
- package/dist/raw/eval/models/EvalExperimentResultMeta.d.ts +102 -0
- package/dist/raw/eval/models/EvalExperimentResultMeta.js +66 -0
- package/dist/raw/eval/models/EvalExperimentStatus.d.ts +31 -0
- package/dist/raw/eval/models/EvalExperimentStatus.js +59 -0
- package/dist/raw/eval/models/EvalGetDataset404Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalGetDataset404Response.js +75 -0
- package/dist/raw/eval/models/EvalGetExperimentTrace400Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalGetExperimentTrace400Response.js +75 -0
- package/dist/raw/eval/models/EvalListDatasets400Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalListDatasets400Response.js +75 -0
- package/dist/raw/eval/models/EvalListDatasets401Response.d.ts +79 -0
- package/dist/raw/eval/models/EvalListDatasets401Response.js +89 -0
- package/dist/raw/eval/models/EvalListDatasets403Response.d.ts +64 -0
- package/dist/raw/eval/models/EvalListDatasets403Response.js +74 -0
- package/dist/raw/eval/models/EvalListDatasets500Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalListDatasets500Response.js +75 -0
- package/dist/raw/eval/models/EvalListDatasets503Response.d.ts +64 -0
- package/dist/raw/eval/models/EvalListDatasets503Response.js +74 -0
- package/dist/raw/eval/models/EvalListDatasets504Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalListDatasets504Response.js +75 -0
- package/dist/raw/eval/models/EvalPageDatasetArtifactView.d.ts +44 -0
- package/dist/raw/eval/models/EvalPageDatasetArtifactView.js +56 -0
- package/dist/raw/eval/models/EvalPageDatasetRecordArtifactView.d.ts +44 -0
- package/dist/raw/eval/models/EvalPageDatasetRecordArtifactView.js +56 -0
- package/dist/raw/eval/models/EvalPageExperimentArtifactView.d.ts +44 -0
- package/dist/raw/eval/models/EvalPageExperimentArtifactView.js +56 -0
- package/dist/raw/eval/models/EvalPageExperimentResultArtifactView.d.ts +44 -0
- package/dist/raw/eval/models/EvalPageExperimentResultArtifactView.js +56 -0
- package/dist/raw/eval/models/EvalPageScorerArtifactView.d.ts +44 -0
- package/dist/raw/eval/models/EvalPageScorerArtifactView.js +56 -0
- package/dist/raw/eval/models/EvalPatchDataset412Response.d.ts +63 -0
- package/dist/raw/eval/models/EvalPatchDataset412Response.js +75 -0
- package/dist/raw/eval/models/EvalRecordSource.d.ts +27 -0
- package/dist/raw/eval/models/EvalRecordSource.js +55 -0
- package/dist/raw/eval/models/EvalScorerArtifactView.d.ts +301 -0
- package/dist/raw/eval/models/EvalScorerArtifactView.js +154 -0
- package/dist/raw/eval/models/EvalScorerCreate.d.ts +172 -0
- package/dist/raw/eval/models/EvalScorerCreate.js +98 -0
- package/dist/raw/eval/models/EvalScorerKind.d.ts +32 -0
- package/dist/raw/eval/models/EvalScorerKind.js +60 -0
- package/dist/raw/eval/models/EvalScorerMeta.d.ts +67 -0
- package/dist/raw/eval/models/EvalScorerMeta.js +60 -0
- package/dist/raw/eval/models/EvalScorerPatch.d.ts +75 -0
- package/dist/raw/eval/models/EvalScorerPatch.js +56 -0
- package/dist/raw/eval/models/EvalScorerScope.d.ts +32 -0
- package/dist/raw/eval/models/EvalScorerScope.js +60 -0
- package/dist/raw/eval/models/EvalSubjectKind.d.ts +33 -0
- package/dist/raw/eval/models/EvalSubjectKind.js +61 -0
- package/dist/raw/eval/models/GatewayArtifactBase.d.ts +250 -0
- package/dist/raw/eval/models/GatewayArtifactBase.js +139 -0
- package/dist/raw/eval/models/GatewayArtifactIdentityRevealView.d.ts +60 -0
- package/dist/raw/eval/models/GatewayArtifactIdentityRevealView.js +58 -0
- package/dist/raw/eval/models/GatewayPageMeta.d.ts +46 -0
- package/dist/raw/eval/models/GatewayPageMeta.js +58 -0
- package/dist/raw/eval/models/GatewaySourceEnvelope.d.ts +106 -0
- package/dist/raw/eval/models/GatewaySourceEnvelope.js +66 -0
- package/dist/raw/eval/models/GatewayTraceStatsBucket.d.ts +25 -0
- package/dist/raw/eval/models/GatewayTraceStatsBucket.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsCompareMode.d.ts +25 -0
- package/dist/raw/eval/models/GatewayTraceStatsCompareMode.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsComparison.d.ts +45 -0
- package/dist/raw/eval/models/GatewayTraceStatsComparison.js +59 -0
- package/dist/raw/eval/models/GatewayTraceStatsData.d.ts +24 -0
- package/dist/raw/eval/models/GatewayTraceStatsData.js +64 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf.d.ts +46 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf.js +66 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf1.d.ts +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf1.js +71 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf2.d.ts +46 -0
- package/dist/raw/eval/models/GatewayTraceStatsDataOneOf2.js +66 -0
- package/dist/raw/eval/models/GatewayTraceStatsEnvelope.d.ts +50 -0
- package/dist/raw/eval/models/GatewayTraceStatsEnvelope.js +59 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilter.d.ts +47 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilter.js +59 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilterOperator.d.ts +28 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilterOperator.js +56 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilterValue.d.ts +21 -0
- package/dist/raw/eval/models/GatewayTraceStatsFilterValue.js +85 -0
- package/dist/raw/eval/models/GatewayTraceStatsGroupBy.d.ts +35 -0
- package/dist/raw/eval/models/GatewayTraceStatsGroupBy.js +63 -0
- package/dist/raw/eval/models/GatewayTraceStatsMeasure.d.ts +39 -0
- package/dist/raw/eval/models/GatewayTraceStatsMeasure.js +67 -0
- package/dist/raw/eval/models/GatewayTraceStatsMeta.d.ts +49 -0
- package/dist/raw/eval/models/GatewayTraceStatsMeta.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrder.d.ts +40 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrder.js +56 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrderBy.d.ts +25 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrderBy.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrderDirection.d.ts +25 -0
- package/dist/raw/eval/models/GatewayTraceStatsOrderDirection.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsPanelQuery.d.ts +69 -0
- package/dist/raw/eval/models/GatewayTraceStatsPanelQuery.js +66 -0
- package/dist/raw/eval/models/GatewayTraceStatsRequest.d.ts +80 -0
- package/dist/raw/eval/models/GatewayTraceStatsRequest.js +72 -0
- package/dist/raw/eval/models/GatewayTraceStatsRequestType.d.ts +26 -0
- package/dist/raw/eval/models/GatewayTraceStatsRequestType.js +54 -0
- package/dist/raw/eval/models/GatewayTraceStatsScalarResult.d.ts +52 -0
- package/dist/raw/eval/models/GatewayTraceStatsScalarResult.js +59 -0
- package/dist/raw/eval/models/GatewayTraceStatsSeries.d.ts +52 -0
- package/dist/raw/eval/models/GatewayTraceStatsSeries.js +62 -0
- package/dist/raw/eval/models/GatewayTraceStatsSeriesPoint.d.ts +38 -0
- package/dist/raw/eval/models/GatewayTraceStatsSeriesPoint.js +52 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableColumn.d.ts +45 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableColumn.js +57 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableColumnType.d.ts +25 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableColumnType.js +53 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableRow.d.ts +38 -0
- package/dist/raw/eval/models/GatewayTraceStatsTableRow.js +54 -0
- package/dist/raw/eval/models/GatewayTraceStatsWindow.d.ts +38 -0
- package/dist/raw/eval/models/GatewayTraceStatsWindow.js +54 -0
- package/dist/raw/eval/models/ProblemDetails.d.ts +76 -0
- package/dist/raw/eval/models/ProblemDetails.js +64 -0
- package/dist/raw/eval/models/RegistryEdgeTarget.d.ts +38 -0
- package/dist/raw/eval/models/RegistryEdgeTarget.js +54 -0
- package/dist/raw/eval/models/RegistryRelationshipEdge.d.ts +96 -0
- package/dist/raw/eval/models/RegistryRelationshipEdge.js +77 -0
- package/dist/raw/eval/models/RegistryRelationshipsView.d.ts +45 -0
- package/dist/raw/eval/models/RegistryRelationshipsView.js +59 -0
- package/dist/raw/eval/models/SpanListEnvelope.d.ts +40 -0
- package/dist/raw/eval/models/SpanListEnvelope.js +56 -0
- package/dist/raw/eval/models/SpanViewDto.d.ts +192 -0
- package/dist/raw/eval/models/SpanViewDto.js +117 -0
- package/dist/raw/eval/models/TraceDetailView.d.ts +177 -0
- package/dist/raw/eval/models/TraceDetailView.js +108 -0
- package/dist/raw/eval/models/TraceLinks.d.ts +45 -0
- package/dist/raw/eval/models/TraceLinks.js +58 -0
- package/dist/raw/eval/models/TraceListEnvelope.d.ts +47 -0
- package/dist/raw/eval/models/TraceListEnvelope.js +61 -0
- package/dist/raw/eval/models/TraceParentView.d.ts +68 -0
- package/dist/raw/eval/models/TraceParentView.js +64 -0
- package/dist/raw/eval/models/TraceStatus.d.ts +26 -0
- package/dist/raw/eval/models/TraceStatus.js +54 -0
- package/dist/raw/eval/models/TraceSummaryStats.d.ts +62 -0
- package/dist/raw/eval/models/TraceSummaryStats.js +62 -0
- package/dist/raw/eval/models/TraceSummaryView.d.ts +150 -0
- package/dist/raw/eval/models/TraceSummaryView.js +96 -0
- package/dist/raw/eval/models/UsageView.d.ts +56 -0
- package/dist/raw/eval/models/UsageView.js +66 -0
- package/dist/raw/eval/models/index.d.ts +100 -0
- package/dist/raw/eval/models/index.js +118 -0
- package/dist/raw/eval/runtime.d.ts +185 -0
- package/dist/raw/eval/runtime.js +369 -0
- package/dist/resources.json +307 -0
- package/dist/tracing/client.js +6 -4
- package/dist/tracing/propagation.d.ts +1 -0
- package/dist/tracing/propagation.js +7 -0
- package/dist/tracing/semconv.d.ts +8 -1
- package/dist/tracing/semconv.js +27 -13
- package/dist/tracing/spans.d.ts +8 -4
- package/dist/tracing/spans.js +28 -13
- package/dist/tracing/version.d.ts +1 -1
- package/dist/tracing/version.js +1 -1
- package/package.json +1 -1
|
@@ -0,0 +1,51 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
/* tslint:disable */
|
|
3
|
+
/* eslint-disable */
|
|
4
|
+
/**
|
|
5
|
+
* Evaluation catalog
|
|
6
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
7
|
+
*
|
|
8
|
+
* The version of the OpenAPI document: 0.1.0
|
|
9
|
+
*
|
|
10
|
+
*
|
|
11
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
12
|
+
* https://openapi-generator.tech
|
|
13
|
+
* Do not edit the class manually.
|
|
14
|
+
*/
|
|
15
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
16
|
+
exports.instanceOfEvalExperimentResultBulkCreate = instanceOfEvalExperimentResultBulkCreate;
|
|
17
|
+
exports.EvalExperimentResultBulkCreateFromJSON = EvalExperimentResultBulkCreateFromJSON;
|
|
18
|
+
exports.EvalExperimentResultBulkCreateFromJSONTyped = EvalExperimentResultBulkCreateFromJSONTyped;
|
|
19
|
+
exports.EvalExperimentResultBulkCreateToJSON = EvalExperimentResultBulkCreateToJSON;
|
|
20
|
+
exports.EvalExperimentResultBulkCreateToJSONTyped = EvalExperimentResultBulkCreateToJSONTyped;
|
|
21
|
+
const EvalExperimentResultCreate_1 = require("./EvalExperimentResultCreate");
|
|
22
|
+
/**
|
|
23
|
+
* Check if a given object implements the EvalExperimentResultBulkCreate interface.
|
|
24
|
+
*/
|
|
25
|
+
function instanceOfEvalExperimentResultBulkCreate(value) {
|
|
26
|
+
if (!('items' in value) || value['items'] === undefined)
|
|
27
|
+
return false;
|
|
28
|
+
return true;
|
|
29
|
+
}
|
|
30
|
+
function EvalExperimentResultBulkCreateFromJSON(json) {
|
|
31
|
+
return EvalExperimentResultBulkCreateFromJSONTyped(json, false);
|
|
32
|
+
}
|
|
33
|
+
function EvalExperimentResultBulkCreateFromJSONTyped(json, ignoreDiscriminator) {
|
|
34
|
+
if (json == null) {
|
|
35
|
+
return json;
|
|
36
|
+
}
|
|
37
|
+
return {
|
|
38
|
+
'items': (json['items'].map(EvalExperimentResultCreate_1.EvalExperimentResultCreateFromJSON)),
|
|
39
|
+
};
|
|
40
|
+
}
|
|
41
|
+
function EvalExperimentResultBulkCreateToJSON(json) {
|
|
42
|
+
return EvalExperimentResultBulkCreateToJSONTyped(json, false);
|
|
43
|
+
}
|
|
44
|
+
function EvalExperimentResultBulkCreateToJSONTyped(value, ignoreDiscriminator = false) {
|
|
45
|
+
if (value == null) {
|
|
46
|
+
return value;
|
|
47
|
+
}
|
|
48
|
+
return {
|
|
49
|
+
'items': (value['items'].map(EvalExperimentResultCreate_1.EvalExperimentResultCreateToJSON)),
|
|
50
|
+
};
|
|
51
|
+
}
|
|
@@ -0,0 +1,197 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Evaluation catalog
|
|
3
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
4
|
+
*
|
|
5
|
+
* The version of the OpenAPI document: 0.1.0
|
|
6
|
+
*
|
|
7
|
+
*
|
|
8
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
9
|
+
* https://openapi-generator.tech
|
|
10
|
+
* Do not edit the class manually.
|
|
11
|
+
*/
|
|
12
|
+
/**
|
|
13
|
+
* The caller-writable surface of one evaluated case.
|
|
14
|
+
*
|
|
15
|
+
* `experiment_id` is not a body field — it comes from the path, since the
|
|
16
|
+
* routes are nested under the owning experiment.
|
|
17
|
+
* @export
|
|
18
|
+
* @interface EvalExperimentResultCreate
|
|
19
|
+
*/
|
|
20
|
+
export interface EvalExperimentResultCreate {
|
|
21
|
+
/**
|
|
22
|
+
* RFC 3339. The upstream's own creation time, never the commit time.
|
|
23
|
+
* @type {string}
|
|
24
|
+
* @memberof EvalExperimentResultCreate
|
|
25
|
+
*/
|
|
26
|
+
sourceCreatedAt?: string;
|
|
27
|
+
/**
|
|
28
|
+
* The upstream author, git `Name <email>` form (distinct from the registry
|
|
29
|
+
* principal `created_by`). The first author; see `source_last_modified_by`
|
|
30
|
+
* for who changed it last.
|
|
31
|
+
* @type {string}
|
|
32
|
+
* @memberof EvalExperimentResultCreate
|
|
33
|
+
*/
|
|
34
|
+
sourceCreatedBy?: string;
|
|
35
|
+
/**
|
|
36
|
+
* What the upstream calls this object. The reconciliation key for a re-sync,
|
|
37
|
+
* and the reason to prefer it over matching on `name`.
|
|
38
|
+
* @type {string}
|
|
39
|
+
* @memberof EvalExperimentResultCreate
|
|
40
|
+
*/
|
|
41
|
+
sourceId?: string;
|
|
42
|
+
/**
|
|
43
|
+
* RFC 3339.
|
|
44
|
+
* @type {string}
|
|
45
|
+
* @memberof EvalExperimentResultCreate
|
|
46
|
+
*/
|
|
47
|
+
sourceLastModifiedAt?: string;
|
|
48
|
+
/**
|
|
49
|
+
* The upstream author of the last change, git `Name <email>` form.
|
|
50
|
+
* @type {string}
|
|
51
|
+
* @memberof EvalExperimentResultCreate
|
|
52
|
+
*/
|
|
53
|
+
sourceLastModifiedBy?: string;
|
|
54
|
+
/**
|
|
55
|
+
*
|
|
56
|
+
* @type {any}
|
|
57
|
+
* @memberof EvalExperimentResultCreate
|
|
58
|
+
*/
|
|
59
|
+
sourceMetadata?: any | null;
|
|
60
|
+
/**
|
|
61
|
+
* RFC 3339. The time of the sync that produced this version.
|
|
62
|
+
* @type {string}
|
|
63
|
+
* @memberof EvalExperimentResultCreate
|
|
64
|
+
*/
|
|
65
|
+
sourceSyncedAt?: string;
|
|
66
|
+
/**
|
|
67
|
+
* The content authority: `git`, `atlan`, `local`, or `external` for a
|
|
68
|
+
* customer's own system. Omitted, the registry defaults it.
|
|
69
|
+
* @type {string}
|
|
70
|
+
* @memberof EvalExperimentResultCreate
|
|
71
|
+
*/
|
|
72
|
+
sourceType?: string;
|
|
73
|
+
/**
|
|
74
|
+
* Scheme-tagged origin — where the upstream object lives.
|
|
75
|
+
* @type {string}
|
|
76
|
+
* @memberof EvalExperimentResultCreate
|
|
77
|
+
*/
|
|
78
|
+
sourceUri?: string;
|
|
79
|
+
/**
|
|
80
|
+
* The upstream revision (a commit hash or tag), distinct from `version_label`.
|
|
81
|
+
* @type {string}
|
|
82
|
+
* @memberof EvalExperimentResultCreate
|
|
83
|
+
*/
|
|
84
|
+
sourceVersion?: string;
|
|
85
|
+
/**
|
|
86
|
+
* Human-readable description.
|
|
87
|
+
* @type {string}
|
|
88
|
+
* @memberof EvalExperimentResultCreate
|
|
89
|
+
*/
|
|
90
|
+
description?: string;
|
|
91
|
+
/**
|
|
92
|
+
* Human-readable label. Omitted, the gateway defaults it to `name`.
|
|
93
|
+
* @type {string}
|
|
94
|
+
* @memberof EvalExperimentResultCreate
|
|
95
|
+
*/
|
|
96
|
+
displayName?: string;
|
|
97
|
+
/**
|
|
98
|
+
* Stable machine handle ([AIP-122](https://google.aip.dev/122)) — short and
|
|
99
|
+
* identifier-shaped, kebab-case by convention. Unique per `(workspace,
|
|
100
|
+
* kind)` among live rows (`artifacts_name_idx`), but not identity: `id`
|
|
101
|
+
* is the only stable handle.
|
|
102
|
+
* @type {string}
|
|
103
|
+
* @memberof EvalExperimentResultCreate
|
|
104
|
+
*/
|
|
105
|
+
name: string;
|
|
106
|
+
/**
|
|
107
|
+
* Optional canonical owner (`users.user_id`). Omitted, the gateway
|
|
108
|
+
* attributes the caller.
|
|
109
|
+
* @type {string}
|
|
110
|
+
* @memberof EvalExperimentResultCreate
|
|
111
|
+
*/
|
|
112
|
+
ownerId?: string;
|
|
113
|
+
/**
|
|
114
|
+
* Optional `project_`-prefixed handle to group under. Must belong to
|
|
115
|
+
* `workspace_id` (else 422).
|
|
116
|
+
* @type {string}
|
|
117
|
+
* @memberof EvalExperimentResultCreate
|
|
118
|
+
*/
|
|
119
|
+
projectId?: string;
|
|
120
|
+
/**
|
|
121
|
+
* Optional `repo_`-prefixed source repository. No default — a repo is a
|
|
122
|
+
* factual claim about provenance, so it is never guessed.
|
|
123
|
+
* @type {string}
|
|
124
|
+
* @memberof EvalExperimentResultCreate
|
|
125
|
+
*/
|
|
126
|
+
repoId?: string;
|
|
127
|
+
/**
|
|
128
|
+
* `workspace_`-prefixed handle to create in. **Required** — there is
|
|
129
|
+
* deliberately no default, because guessing a tenant's namespace is
|
|
130
|
+
* invisible until someone else reads the data.
|
|
131
|
+
* @type {string}
|
|
132
|
+
* @memberof EvalExperimentResultCreate
|
|
133
|
+
*/
|
|
134
|
+
workspaceId: string;
|
|
135
|
+
/**
|
|
136
|
+
* The dataset row this case came from; omit for an ad-hoc case.
|
|
137
|
+
* @type {string}
|
|
138
|
+
* @memberof EvalExperimentResultCreate
|
|
139
|
+
*/
|
|
140
|
+
datasetRecordId?: string | null;
|
|
141
|
+
/**
|
|
142
|
+
* Wall-clock duration of the case, milliseconds.
|
|
143
|
+
* @type {number}
|
|
144
|
+
* @memberof EvalExperimentResultCreate
|
|
145
|
+
*/
|
|
146
|
+
durationMs?: number | null;
|
|
147
|
+
/**
|
|
148
|
+
* The failure, when the case errored instead of scoring.
|
|
149
|
+
* @type {string}
|
|
150
|
+
* @memberof EvalExperimentResultCreate
|
|
151
|
+
*/
|
|
152
|
+
error?: string | null;
|
|
153
|
+
/**
|
|
154
|
+
* The expected outcome as run, snapshot at run time.
|
|
155
|
+
* @type {{ [key: string]: any; }}
|
|
156
|
+
* @memberof EvalExperimentResultCreate
|
|
157
|
+
*/
|
|
158
|
+
expected?: {
|
|
159
|
+
[key: string]: any;
|
|
160
|
+
} | null;
|
|
161
|
+
/**
|
|
162
|
+
* The case's input as run, snapshot at run time.
|
|
163
|
+
* @type {{ [key: string]: any; }}
|
|
164
|
+
* @memberof EvalExperimentResultCreate
|
|
165
|
+
*/
|
|
166
|
+
input?: {
|
|
167
|
+
[key: string]: any;
|
|
168
|
+
} | null;
|
|
169
|
+
/**
|
|
170
|
+
* What the subject produced.
|
|
171
|
+
* @type {{ [key: string]: any; }}
|
|
172
|
+
* @memberof EvalExperimentResultCreate
|
|
173
|
+
*/
|
|
174
|
+
output?: {
|
|
175
|
+
[key: string]: any;
|
|
176
|
+
} | null;
|
|
177
|
+
/**
|
|
178
|
+
* The gateway session this case ran as, when the run was cataloged.
|
|
179
|
+
* @type {string}
|
|
180
|
+
* @memberof EvalExperimentResultCreate
|
|
181
|
+
*/
|
|
182
|
+
sessionId?: string | null;
|
|
183
|
+
/**
|
|
184
|
+
* The OTel trace id of the case's execution.
|
|
185
|
+
* @type {string}
|
|
186
|
+
* @memberof EvalExperimentResultCreate
|
|
187
|
+
*/
|
|
188
|
+
traceId?: string | null;
|
|
189
|
+
}
|
|
190
|
+
/**
|
|
191
|
+
* Check if a given object implements the EvalExperimentResultCreate interface.
|
|
192
|
+
*/
|
|
193
|
+
export declare function instanceOfEvalExperimentResultCreate(value: object): value is EvalExperimentResultCreate;
|
|
194
|
+
export declare function EvalExperimentResultCreateFromJSON(json: any): EvalExperimentResultCreate;
|
|
195
|
+
export declare function EvalExperimentResultCreateFromJSONTyped(json: any, ignoreDiscriminator: boolean): EvalExperimentResultCreate;
|
|
196
|
+
export declare function EvalExperimentResultCreateToJSON(json: any): EvalExperimentResultCreate;
|
|
197
|
+
export declare function EvalExperimentResultCreateToJSONTyped(value?: EvalExperimentResultCreate | null, ignoreDiscriminator?: boolean): any;
|
|
@@ -0,0 +1,100 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
/* tslint:disable */
|
|
3
|
+
/* eslint-disable */
|
|
4
|
+
/**
|
|
5
|
+
* Evaluation catalog
|
|
6
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
7
|
+
*
|
|
8
|
+
* The version of the OpenAPI document: 0.1.0
|
|
9
|
+
*
|
|
10
|
+
*
|
|
11
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
12
|
+
* https://openapi-generator.tech
|
|
13
|
+
* Do not edit the class manually.
|
|
14
|
+
*/
|
|
15
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
16
|
+
exports.instanceOfEvalExperimentResultCreate = instanceOfEvalExperimentResultCreate;
|
|
17
|
+
exports.EvalExperimentResultCreateFromJSON = EvalExperimentResultCreateFromJSON;
|
|
18
|
+
exports.EvalExperimentResultCreateFromJSONTyped = EvalExperimentResultCreateFromJSONTyped;
|
|
19
|
+
exports.EvalExperimentResultCreateToJSON = EvalExperimentResultCreateToJSON;
|
|
20
|
+
exports.EvalExperimentResultCreateToJSONTyped = EvalExperimentResultCreateToJSONTyped;
|
|
21
|
+
/**
|
|
22
|
+
* Check if a given object implements the EvalExperimentResultCreate interface.
|
|
23
|
+
*/
|
|
24
|
+
function instanceOfEvalExperimentResultCreate(value) {
|
|
25
|
+
if (!('name' in value) || value['name'] === undefined)
|
|
26
|
+
return false;
|
|
27
|
+
if ((!('workspaceId' in value) && !('workspace_id' in value)) || (value['workspaceId'] === undefined && value['workspace_id'] === undefined))
|
|
28
|
+
return false;
|
|
29
|
+
return true;
|
|
30
|
+
}
|
|
31
|
+
function EvalExperimentResultCreateFromJSON(json) {
|
|
32
|
+
return EvalExperimentResultCreateFromJSONTyped(json, false);
|
|
33
|
+
}
|
|
34
|
+
function EvalExperimentResultCreateFromJSONTyped(json, ignoreDiscriminator) {
|
|
35
|
+
if (json == null) {
|
|
36
|
+
return json;
|
|
37
|
+
}
|
|
38
|
+
return {
|
|
39
|
+
'sourceCreatedAt': json['source_created_at'] == null ? undefined : json['source_created_at'],
|
|
40
|
+
'sourceCreatedBy': json['source_created_by'] == null ? undefined : json['source_created_by'],
|
|
41
|
+
'sourceId': json['source_id'] == null ? undefined : json['source_id'],
|
|
42
|
+
'sourceLastModifiedAt': json['source_last_modified_at'] == null ? undefined : json['source_last_modified_at'],
|
|
43
|
+
'sourceLastModifiedBy': json['source_last_modified_by'] == null ? undefined : json['source_last_modified_by'],
|
|
44
|
+
'sourceMetadata': json['source_metadata'] === undefined ? undefined : json['source_metadata'] === null ? null : json['source_metadata'],
|
|
45
|
+
'sourceSyncedAt': json['source_synced_at'] == null ? undefined : json['source_synced_at'],
|
|
46
|
+
'sourceType': json['source_type'] == null ? undefined : json['source_type'],
|
|
47
|
+
'sourceUri': json['source_uri'] == null ? undefined : json['source_uri'],
|
|
48
|
+
'sourceVersion': json['source_version'] == null ? undefined : json['source_version'],
|
|
49
|
+
'description': json['description'] == null ? undefined : json['description'],
|
|
50
|
+
'displayName': json['display_name'] == null ? undefined : json['display_name'],
|
|
51
|
+
'name': json['name'],
|
|
52
|
+
'ownerId': json['owner_id'] == null ? undefined : json['owner_id'],
|
|
53
|
+
'projectId': json['project_id'] == null ? undefined : json['project_id'],
|
|
54
|
+
'repoId': json['repo_id'] == null ? undefined : json['repo_id'],
|
|
55
|
+
'workspaceId': json['workspace_id'],
|
|
56
|
+
'datasetRecordId': json['dataset_record_id'] === undefined ? undefined : json['dataset_record_id'] === null ? null : json['dataset_record_id'],
|
|
57
|
+
'durationMs': json['duration_ms'] === undefined ? undefined : json['duration_ms'] === null ? null : json['duration_ms'],
|
|
58
|
+
'error': json['error'] === undefined ? undefined : json['error'] === null ? null : json['error'],
|
|
59
|
+
'expected': json['expected'] === undefined ? undefined : json['expected'] === null ? null : json['expected'],
|
|
60
|
+
'input': json['input'] === undefined ? undefined : json['input'] === null ? null : json['input'],
|
|
61
|
+
'output': json['output'] === undefined ? undefined : json['output'] === null ? null : json['output'],
|
|
62
|
+
'sessionId': json['session_id'] === undefined ? undefined : json['session_id'] === null ? null : json['session_id'],
|
|
63
|
+
'traceId': json['trace_id'] === undefined ? undefined : json['trace_id'] === null ? null : json['trace_id'],
|
|
64
|
+
};
|
|
65
|
+
}
|
|
66
|
+
function EvalExperimentResultCreateToJSON(json) {
|
|
67
|
+
return EvalExperimentResultCreateToJSONTyped(json, false);
|
|
68
|
+
}
|
|
69
|
+
function EvalExperimentResultCreateToJSONTyped(value, ignoreDiscriminator = false) {
|
|
70
|
+
if (value == null) {
|
|
71
|
+
return value;
|
|
72
|
+
}
|
|
73
|
+
return {
|
|
74
|
+
'source_created_at': value['sourceCreatedAt'],
|
|
75
|
+
'source_created_by': value['sourceCreatedBy'],
|
|
76
|
+
'source_id': value['sourceId'],
|
|
77
|
+
'source_last_modified_at': value['sourceLastModifiedAt'],
|
|
78
|
+
'source_last_modified_by': value['sourceLastModifiedBy'],
|
|
79
|
+
'source_metadata': value['sourceMetadata'],
|
|
80
|
+
'source_synced_at': value['sourceSyncedAt'],
|
|
81
|
+
'source_type': value['sourceType'],
|
|
82
|
+
'source_uri': value['sourceUri'],
|
|
83
|
+
'source_version': value['sourceVersion'],
|
|
84
|
+
'description': value['description'],
|
|
85
|
+
'display_name': value['displayName'],
|
|
86
|
+
'name': value['name'],
|
|
87
|
+
'owner_id': value['ownerId'],
|
|
88
|
+
'project_id': value['projectId'],
|
|
89
|
+
'repo_id': value['repoId'],
|
|
90
|
+
'workspace_id': value['workspaceId'],
|
|
91
|
+
'dataset_record_id': value['datasetRecordId'],
|
|
92
|
+
'duration_ms': value['durationMs'],
|
|
93
|
+
'error': value['error'],
|
|
94
|
+
'expected': value['expected'],
|
|
95
|
+
'input': value['input'],
|
|
96
|
+
'output': value['output'],
|
|
97
|
+
'session_id': value['sessionId'],
|
|
98
|
+
'trace_id': value['traceId'],
|
|
99
|
+
};
|
|
100
|
+
}
|
|
@@ -0,0 +1,102 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Evaluation catalog
|
|
3
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
4
|
+
*
|
|
5
|
+
* The version of the OpenAPI document: 0.1.0
|
|
6
|
+
*
|
|
7
|
+
*
|
|
8
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
9
|
+
* https://openapi-generator.tech
|
|
10
|
+
* Do not edit the class manually.
|
|
11
|
+
*/
|
|
12
|
+
/**
|
|
13
|
+
* The `experiment_result` artifact kind: one evaluated case of an experiment.
|
|
14
|
+
*
|
|
15
|
+
* **Create-only**, for the same reason `session_message` is
|
|
16
|
+
* (artifacts-2026-08-18-create-only-kinds): a result is reported after the
|
|
17
|
+
* fact, so the upload is the whole write, and an editable result would make a
|
|
18
|
+
* completed experiment's scores rewritable history.
|
|
19
|
+
*
|
|
20
|
+
* `input`/`expected` are snapshot copies taken at run time — the dataset
|
|
21
|
+
* record may be curated afterwards, and the result must keep describing the
|
|
22
|
+
* case that actually ran. `session_id` and `trace_id` are join keys into the
|
|
23
|
+
* run's transcript and its OTel trace; both optional, both the same seams the
|
|
24
|
+
* agent extension already reads.
|
|
25
|
+
* @export
|
|
26
|
+
* @interface EvalExperimentResultMeta
|
|
27
|
+
*/
|
|
28
|
+
export interface EvalExperimentResultMeta {
|
|
29
|
+
/**
|
|
30
|
+
* The dataset row this case came from; absent for an ad-hoc case. The
|
|
31
|
+
* per-case join key when comparing two experiments over one dataset.
|
|
32
|
+
* @type {string}
|
|
33
|
+
* @memberof EvalExperimentResultMeta
|
|
34
|
+
*/
|
|
35
|
+
datasetRecordId?: string | null;
|
|
36
|
+
/**
|
|
37
|
+
* Wall-clock duration of the case, milliseconds.
|
|
38
|
+
* @type {number}
|
|
39
|
+
* @memberof EvalExperimentResultMeta
|
|
40
|
+
*/
|
|
41
|
+
durationMs?: number | null;
|
|
42
|
+
/**
|
|
43
|
+
* The failure, when the case errored instead of scoring. Whatever the
|
|
44
|
+
* runner reported, including a stack trace.
|
|
45
|
+
* @type {string}
|
|
46
|
+
* @memberof EvalExperimentResultMeta
|
|
47
|
+
*/
|
|
48
|
+
error?: string | null;
|
|
49
|
+
/**
|
|
50
|
+
* The expected outcome as run, snapshot at run time.
|
|
51
|
+
* @type {{ [key: string]: any; }}
|
|
52
|
+
* @memberof EvalExperimentResultMeta
|
|
53
|
+
*/
|
|
54
|
+
expected?: {
|
|
55
|
+
[key: string]: any;
|
|
56
|
+
} | null;
|
|
57
|
+
/**
|
|
58
|
+
* The owning `experiment` handle. Creates are refused once it is terminal.
|
|
59
|
+
* @type {string}
|
|
60
|
+
* @memberof EvalExperimentResultMeta
|
|
61
|
+
*/
|
|
62
|
+
experimentId: string;
|
|
63
|
+
/**
|
|
64
|
+
* The case's input as run, snapshot at run time.
|
|
65
|
+
* @type {{ [key: string]: any; }}
|
|
66
|
+
* @memberof EvalExperimentResultMeta
|
|
67
|
+
*/
|
|
68
|
+
input?: {
|
|
69
|
+
[key: string]: any;
|
|
70
|
+
} | null;
|
|
71
|
+
/**
|
|
72
|
+
* What the subject produced.
|
|
73
|
+
* @type {{ [key: string]: any; }}
|
|
74
|
+
* @memberof EvalExperimentResultMeta
|
|
75
|
+
*/
|
|
76
|
+
output?: {
|
|
77
|
+
[key: string]: any;
|
|
78
|
+
} | null;
|
|
79
|
+
/**
|
|
80
|
+
* The gateway `session` this case ran as, when the run was cataloged
|
|
81
|
+
* (resolved at create). Links a result to its transcript.
|
|
82
|
+
* @type {string}
|
|
83
|
+
* @memberof EvalExperimentResultMeta
|
|
84
|
+
*/
|
|
85
|
+
sessionId?: string | null;
|
|
86
|
+
/**
|
|
87
|
+
* The OTel trace id of the case's execution — the same join seam
|
|
88
|
+
* `session.external_session_id` uses. Recorded verbatim, never resolved
|
|
89
|
+
* (a trace is not a registry artifact).
|
|
90
|
+
* @type {string}
|
|
91
|
+
* @memberof EvalExperimentResultMeta
|
|
92
|
+
*/
|
|
93
|
+
traceId?: string | null;
|
|
94
|
+
}
|
|
95
|
+
/**
|
|
96
|
+
* Check if a given object implements the EvalExperimentResultMeta interface.
|
|
97
|
+
*/
|
|
98
|
+
export declare function instanceOfEvalExperimentResultMeta(value: object): value is EvalExperimentResultMeta;
|
|
99
|
+
export declare function EvalExperimentResultMetaFromJSON(json: any): EvalExperimentResultMeta;
|
|
100
|
+
export declare function EvalExperimentResultMetaFromJSONTyped(json: any, ignoreDiscriminator: boolean): EvalExperimentResultMeta;
|
|
101
|
+
export declare function EvalExperimentResultMetaToJSON(json: any): EvalExperimentResultMeta;
|
|
102
|
+
export declare function EvalExperimentResultMetaToJSONTyped(value?: EvalExperimentResultMeta | null, ignoreDiscriminator?: boolean): any;
|
|
@@ -0,0 +1,66 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
/* tslint:disable */
|
|
3
|
+
/* eslint-disable */
|
|
4
|
+
/**
|
|
5
|
+
* Evaluation catalog
|
|
6
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
7
|
+
*
|
|
8
|
+
* The version of the OpenAPI document: 0.1.0
|
|
9
|
+
*
|
|
10
|
+
*
|
|
11
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
12
|
+
* https://openapi-generator.tech
|
|
13
|
+
* Do not edit the class manually.
|
|
14
|
+
*/
|
|
15
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
16
|
+
exports.instanceOfEvalExperimentResultMeta = instanceOfEvalExperimentResultMeta;
|
|
17
|
+
exports.EvalExperimentResultMetaFromJSON = EvalExperimentResultMetaFromJSON;
|
|
18
|
+
exports.EvalExperimentResultMetaFromJSONTyped = EvalExperimentResultMetaFromJSONTyped;
|
|
19
|
+
exports.EvalExperimentResultMetaToJSON = EvalExperimentResultMetaToJSON;
|
|
20
|
+
exports.EvalExperimentResultMetaToJSONTyped = EvalExperimentResultMetaToJSONTyped;
|
|
21
|
+
/**
|
|
22
|
+
* Check if a given object implements the EvalExperimentResultMeta interface.
|
|
23
|
+
*/
|
|
24
|
+
function instanceOfEvalExperimentResultMeta(value) {
|
|
25
|
+
if ((!('experimentId' in value) && !('experiment_id' in value)) || (value['experimentId'] === undefined && value['experiment_id'] === undefined))
|
|
26
|
+
return false;
|
|
27
|
+
return true;
|
|
28
|
+
}
|
|
29
|
+
function EvalExperimentResultMetaFromJSON(json) {
|
|
30
|
+
return EvalExperimentResultMetaFromJSONTyped(json, false);
|
|
31
|
+
}
|
|
32
|
+
function EvalExperimentResultMetaFromJSONTyped(json, ignoreDiscriminator) {
|
|
33
|
+
if (json == null) {
|
|
34
|
+
return json;
|
|
35
|
+
}
|
|
36
|
+
return {
|
|
37
|
+
'datasetRecordId': json['dataset_record_id'] === undefined ? undefined : json['dataset_record_id'] === null ? null : json['dataset_record_id'],
|
|
38
|
+
'durationMs': json['duration_ms'] === undefined ? undefined : json['duration_ms'] === null ? null : json['duration_ms'],
|
|
39
|
+
'error': json['error'] === undefined ? undefined : json['error'] === null ? null : json['error'],
|
|
40
|
+
'expected': json['expected'] === undefined ? undefined : json['expected'] === null ? null : json['expected'],
|
|
41
|
+
'experimentId': json['experiment_id'],
|
|
42
|
+
'input': json['input'] === undefined ? undefined : json['input'] === null ? null : json['input'],
|
|
43
|
+
'output': json['output'] === undefined ? undefined : json['output'] === null ? null : json['output'],
|
|
44
|
+
'sessionId': json['session_id'] === undefined ? undefined : json['session_id'] === null ? null : json['session_id'],
|
|
45
|
+
'traceId': json['trace_id'] === undefined ? undefined : json['trace_id'] === null ? null : json['trace_id'],
|
|
46
|
+
};
|
|
47
|
+
}
|
|
48
|
+
function EvalExperimentResultMetaToJSON(json) {
|
|
49
|
+
return EvalExperimentResultMetaToJSONTyped(json, false);
|
|
50
|
+
}
|
|
51
|
+
function EvalExperimentResultMetaToJSONTyped(value, ignoreDiscriminator = false) {
|
|
52
|
+
if (value == null) {
|
|
53
|
+
return value;
|
|
54
|
+
}
|
|
55
|
+
return {
|
|
56
|
+
'dataset_record_id': value['datasetRecordId'],
|
|
57
|
+
'duration_ms': value['durationMs'],
|
|
58
|
+
'error': value['error'],
|
|
59
|
+
'expected': value['expected'],
|
|
60
|
+
'experiment_id': value['experimentId'],
|
|
61
|
+
'input': value['input'],
|
|
62
|
+
'output': value['output'],
|
|
63
|
+
'session_id': value['sessionId'],
|
|
64
|
+
'trace_id': value['traceId'],
|
|
65
|
+
};
|
|
66
|
+
}
|
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Evaluation catalog
|
|
3
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
4
|
+
*
|
|
5
|
+
* The version of the OpenAPI document: 0.1.0
|
|
6
|
+
*
|
|
7
|
+
*
|
|
8
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
9
|
+
* https://openapi-generator.tech
|
|
10
|
+
* Do not edit the class manually.
|
|
11
|
+
*/
|
|
12
|
+
/**
|
|
13
|
+
* Where an experiment stands. Promoted to the indexed `experiment_status`
|
|
14
|
+
* column; doc-comment-free per variant.
|
|
15
|
+
*
|
|
16
|
+
* `completed` and `failed` are terminal: `update.pre` freezes a terminal
|
|
17
|
+
* experiment and `experiment_result` creates are refused against one, which is
|
|
18
|
+
* what makes a finished experiment an immutable snapshot.
|
|
19
|
+
* @export
|
|
20
|
+
* @enum {string}
|
|
21
|
+
*/
|
|
22
|
+
export declare enum EvalExperimentStatus {
|
|
23
|
+
Running = "running",
|
|
24
|
+
Completed = "completed",
|
|
25
|
+
Failed = "failed"
|
|
26
|
+
}
|
|
27
|
+
export declare function instanceOfEvalExperimentStatus(value: any): boolean;
|
|
28
|
+
export declare function EvalExperimentStatusFromJSON(json: any): EvalExperimentStatus;
|
|
29
|
+
export declare function EvalExperimentStatusFromJSONTyped(json: any, ignoreDiscriminator: boolean): EvalExperimentStatus;
|
|
30
|
+
export declare function EvalExperimentStatusToJSON(value?: EvalExperimentStatus | null): any;
|
|
31
|
+
export declare function EvalExperimentStatusToJSONTyped(value: any, ignoreDiscriminator: boolean): EvalExperimentStatus;
|
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
/* tslint:disable */
|
|
3
|
+
/* eslint-disable */
|
|
4
|
+
/**
|
|
5
|
+
* Evaluation catalog
|
|
6
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
7
|
+
*
|
|
8
|
+
* The version of the OpenAPI document: 0.1.0
|
|
9
|
+
*
|
|
10
|
+
*
|
|
11
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
12
|
+
* https://openapi-generator.tech
|
|
13
|
+
* Do not edit the class manually.
|
|
14
|
+
*/
|
|
15
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
16
|
+
exports.EvalExperimentStatus = void 0;
|
|
17
|
+
exports.instanceOfEvalExperimentStatus = instanceOfEvalExperimentStatus;
|
|
18
|
+
exports.EvalExperimentStatusFromJSON = EvalExperimentStatusFromJSON;
|
|
19
|
+
exports.EvalExperimentStatusFromJSONTyped = EvalExperimentStatusFromJSONTyped;
|
|
20
|
+
exports.EvalExperimentStatusToJSON = EvalExperimentStatusToJSON;
|
|
21
|
+
exports.EvalExperimentStatusToJSONTyped = EvalExperimentStatusToJSONTyped;
|
|
22
|
+
/**
|
|
23
|
+
* Where an experiment stands. Promoted to the indexed `experiment_status`
|
|
24
|
+
* column; doc-comment-free per variant.
|
|
25
|
+
*
|
|
26
|
+
* `completed` and `failed` are terminal: `update.pre` freezes a terminal
|
|
27
|
+
* experiment and `experiment_result` creates are refused against one, which is
|
|
28
|
+
* what makes a finished experiment an immutable snapshot.
|
|
29
|
+
* @export
|
|
30
|
+
* @enum {string}
|
|
31
|
+
*/
|
|
32
|
+
var EvalExperimentStatus;
|
|
33
|
+
(function (EvalExperimentStatus) {
|
|
34
|
+
EvalExperimentStatus["Running"] = "running";
|
|
35
|
+
EvalExperimentStatus["Completed"] = "completed";
|
|
36
|
+
EvalExperimentStatus["Failed"] = "failed";
|
|
37
|
+
})(EvalExperimentStatus || (exports.EvalExperimentStatus = EvalExperimentStatus = {}));
|
|
38
|
+
function instanceOfEvalExperimentStatus(value) {
|
|
39
|
+
for (const key in EvalExperimentStatus) {
|
|
40
|
+
if (Object.prototype.hasOwnProperty.call(EvalExperimentStatus, key)) {
|
|
41
|
+
if (EvalExperimentStatus[key] === value) {
|
|
42
|
+
return true;
|
|
43
|
+
}
|
|
44
|
+
}
|
|
45
|
+
}
|
|
46
|
+
return false;
|
|
47
|
+
}
|
|
48
|
+
function EvalExperimentStatusFromJSON(json) {
|
|
49
|
+
return EvalExperimentStatusFromJSONTyped(json, false);
|
|
50
|
+
}
|
|
51
|
+
function EvalExperimentStatusFromJSONTyped(json, ignoreDiscriminator) {
|
|
52
|
+
return json;
|
|
53
|
+
}
|
|
54
|
+
function EvalExperimentStatusToJSON(value) {
|
|
55
|
+
return value;
|
|
56
|
+
}
|
|
57
|
+
function EvalExperimentStatusToJSONTyped(value, ignoreDiscriminator) {
|
|
58
|
+
return value;
|
|
59
|
+
}
|
|
@@ -0,0 +1,63 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Evaluation catalog
|
|
3
|
+
* The evaluation loop as registry artifacts: curated `dataset`s of examples pinned from production sessions and traces, `experiment` run snapshots with create-only per-case `experiment_result`s, and versioned `scorer` definitions. A verdict is a score span in the run\'s trace; `POST /experiments/{id}/summarize` rolls a run\'s score spans into its durable `summary`. The gateway executes and scores nothing — external harnesses, CI, and judges run the evals, emit score spans, and upload results here.
|
|
4
|
+
*
|
|
5
|
+
* The version of the OpenAPI document: 0.1.0
|
|
6
|
+
*
|
|
7
|
+
*
|
|
8
|
+
* NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
|
|
9
|
+
* https://openapi-generator.tech
|
|
10
|
+
* Do not edit the class manually.
|
|
11
|
+
*/
|
|
12
|
+
/**
|
|
13
|
+
* RFC 7807 problem document for 404 (Not Found).
|
|
14
|
+
* @export
|
|
15
|
+
* @interface EvalGetDataset404Response
|
|
16
|
+
*/
|
|
17
|
+
export interface EvalGetDataset404Response {
|
|
18
|
+
/**
|
|
19
|
+
* Stable machine-readable reason for this failure.
|
|
20
|
+
* @type {EvalGetDataset404ResponseCodeEnum}
|
|
21
|
+
* @memberof EvalGetDataset404Response
|
|
22
|
+
*/
|
|
23
|
+
code: EvalGetDataset404ResponseCodeEnum;
|
|
24
|
+
/**
|
|
25
|
+
* Value of the `detail` field.
|
|
26
|
+
* @type {string}
|
|
27
|
+
* @memberof EvalGetDataset404Response
|
|
28
|
+
*/
|
|
29
|
+
detail: string;
|
|
30
|
+
/**
|
|
31
|
+
* Current lifecycle state.
|
|
32
|
+
* @type {number}
|
|
33
|
+
* @memberof EvalGetDataset404Response
|
|
34
|
+
*/
|
|
35
|
+
status: number;
|
|
36
|
+
/**
|
|
37
|
+
* Value of the `title` field.
|
|
38
|
+
* @type {string}
|
|
39
|
+
* @memberof EvalGetDataset404Response
|
|
40
|
+
*/
|
|
41
|
+
title: string;
|
|
42
|
+
/**
|
|
43
|
+
* Stable identifier of the related trace resource.
|
|
44
|
+
* @type {string}
|
|
45
|
+
* @memberof EvalGetDataset404Response
|
|
46
|
+
*/
|
|
47
|
+
traceId?: string;
|
|
48
|
+
}
|
|
49
|
+
/**
|
|
50
|
+
* @export
|
|
51
|
+
* @enum {string}
|
|
52
|
+
*/
|
|
53
|
+
export declare enum EvalGetDataset404ResponseCodeEnum {
|
|
54
|
+
NotFound = "not_found"
|
|
55
|
+
}
|
|
56
|
+
/**
|
|
57
|
+
* Check if a given object implements the EvalGetDataset404Response interface.
|
|
58
|
+
*/
|
|
59
|
+
export declare function instanceOfEvalGetDataset404Response(value: object): value is EvalGetDataset404Response;
|
|
60
|
+
export declare function EvalGetDataset404ResponseFromJSON(json: any): EvalGetDataset404Response;
|
|
61
|
+
export declare function EvalGetDataset404ResponseFromJSONTyped(json: any, ignoreDiscriminator: boolean): EvalGetDataset404Response;
|
|
62
|
+
export declare function EvalGetDataset404ResponseToJSON(json: any): EvalGetDataset404Response;
|
|
63
|
+
export declare function EvalGetDataset404ResponseToJSONTyped(value?: EvalGetDataset404Response | null, ignoreDiscriminator?: boolean): any;
|