uploadArtifact
Upload an artifact file to this workspace. Behavior depends on file.fileType:
- CSV: the file head is parsed to validate encoding/delimiter/columns
and a header-only
dataSchemais synthesized with every column typedString. The artifact is persisted withuploadStatus = READY,schemaInferenceStatus = NOT_REQUESTED, and a non-nulldataSchema. To refine the all-Stringtypes into real ones, callers invokerunArtifactSchemaInferenceas an explicit follow-up step. - XLSX / XLS / NUMBERS / ODS: the file is stored as-is with no
upload-time parsing. The artifact is persisted with
uploadStatus = READY,schemaInferenceStatus = NOT_REQUESTED, anddataSchema = null. Parsing support for these formats is owned by a sibling workstream. - ZIP: the archive is stored as-is with no extraction, schema inference,
data preview, publish, or ingestion work. The artifact is persisted with
uploadStatus = READY,schemaInferenceStatus = NOT_REQUESTED, anddataSchema = null. Schema inference is never run during upload. TheinferenceandinferenceErrorfields on the returnedArtifactUploadResultare always null here — they are populated only byrunArtifactSchemaInference, which reuses the same result shape.
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Headers
Workspace slug. Overrides subdomain-based workspace resolution when provided.
"acme"
Optional request correlation ID echoed back in the response.
"550e8400-e29b-41d4-a716-446655440000"
Body
Response
Successful response
Shared result shape for uploadArtifact and runArtifactSchemaInference.
Carries the persisted artifact record alongside optional inference output
and a small sample-row preview so the schema editor can render without an
extra round trip.
Inference fields are populated only by runArtifactSchemaInference. In
uploadArtifact responses, inference and inferenceError are always
null (upload never runs inference).