Skip to content

api

API Endpoint Reference

Locate canonical organization-scoped operations and their request/response schemas.

API Endpoint Reference

Definition

Use this catalog to locate an operation. The OpenAPI YAML and JSON supply its full body, query, response and security schemas. Authentication is operation-specific; native CLI delegation does not authorize every resource below.

Projects

These paths follow /v1/organizations/{organizationKey}:

MethodPathOperation
GET/projectsList Projects
POST/projectsCreate a Project
GET/projects/{projectId}Read a Project
PATCH/projects/{projectId}Update a Project

Dimensions and terms

These and the following project operations follow /v1/organizations/{organizationKey}/projects/{projectId}. Keep the API's dimensions/terms wire names when working with the corresponding product coverage concepts.

MethodPathOperation
GET, POST/dimensionsList/create dimensions
GET, PATCH/dimensions/{dimensionId}Read/update a dimension
GET, POST/dimensions/{dimensionId}/termsList/create terms
GET, PATCH/dimensions/{dimensionId}/terms/{termId}Read/update a term

Dataset records and views

MethodPathOperation
GET/datasets/viewsList views
GET/datasets/views/{viewId}Read a view
GET/datasets/views/{viewId}/recordsRead view records
GET/datasets/views/{viewId}/representationRead representation
GET/datasets/recordsList records
GET/datasets/records/{recordId}Read a record
GET/datasets/records/{recordId}/versionsList record versions
POST/datasets/uploadUpload records using its operation schema

Cases and Fieldloop

MethodPathGuide
GET/cases, /cases/{caseId}Case reads
PUT/cases/{caseId}/reference-resultReference responses
POST/cases/reference-resultsBatch reference responses
POST/cases/importsMultipart imports
GET/cases/imports/{importId}Import status
POST/fieldloop/sources/{sourceId}/batchesFieldloop submission
GET/fieldloop/sources/{sourceId}/batches/{batchId}Fieldloop receipt

Case imports and batch reference writes reject repeated idempotency keys rather than replaying receipts. Use their specific recovery rules.

Benchmark context and external outputs

The following paths continue beneath /benchmarks/{benchmarkId}/versions/{benchmarkVersionId}. run-groups identifies the product Run; a nested child runId identifies a Trial.

MethodPathOperation
GET/version-contextPinned summary/Cases/Policies/Rubrics
GET/case-manifestFrozen execution population
GET/run-groupsExisting groups and submission context
POST/run-groupsCreate a group/first child with actual results
POST/run-groups/{groupId}/runsCreate an additional Trial
GET/run-groups/{groupId}/runs/{runId}Import state
GET, POST/run-groups/{groupId}/runs/{runId}/resultsRead/append submitted outputs
POST/run-groups/{groupId}/runs/{runId}/completeFreeze outputs and request evaluation

Follow External Evaluation Outputs for the complete lifecycle and immutable-response rules.

Managed execution and assessment

MethodPathOperation
GET/managed-run-profileCurrent execution policy/token
POST/harness-submissionsSave exact source
POST/harness-runtime-preparationsRequest runtime preparation
POST/managed-runsRequest managed execution
GET/runs/{runId}Individual execution/progress
GET/runs/{runId}/evaluation-contextBounded context and observations
GET/runs/{runId}/evaluation-resultsPaged judgments
GET/runs/{runId}/evaluation-instructionsEvaluation procedure
GET/comparison/{runId}/summaryExplicit baseline/target summary
GET/comparison/{runId}/casesPaged Case changes

See Managed Execution and Results for sequence, filters and interpretation.

Found something unclear?

Report outdated, unsupported, or confusing docs so we can fix the source page.

Report a docs issue

Continue learning

Related docs