Datasets

DatasetsClient

adaline.datasets manages datasets as a whole — create a dataset, look up metadata and columns, rename, or delete. Row-level and column-level operations live on the nested .rows and .columns sub-clients.

Access

import { Adaline } from '@adaline/client';

const adaline = new Adaline();
const datasets = adaline.datasets; // DatasetsClient

The class is also exported directly for dependency injection or testing:

import { DatasetsClient } from '@adaline/client';

Sub-clients

DatasetsClient exposes two nested namespaces:

Property	Client	Covers
adaline.datasets.rows	DatasetRowsClient	List / create / update / delete dataset rows
adaline.datasets.columns	DatasetColumnsClient	Create / update / delete columns + fetchDynamic to resolve dynamic columns

Types used below come from @adaline/api:

import type {
  Dataset,
  DatasetSummary,
  CreateDatasetRequest,
  UpdateDatasetRequest,
  ListDatasetsResponse,
  SortOrder,
} from '@adaline/api';

---

list()

List datasets in a project (paginated).

list(options: {
  projectId: string;
  sort?: SortOrder;
  createdAfter?: number;
  createdBefore?: number;
  limit?: number;
  cursor?: string;
}): Promise<ListDatasetsResponse>

Parameters

Name	Type	Required	Description
projectId	string	Yes	Project whose datasets should be returned.
sort	SortOrder	No	"createdAt:asc" or "createdAt:desc".
createdAfter	number	No	Unix milliseconds. Keep only datasets created after this instant.
createdBefore	number	No	Unix milliseconds. Keep only datasets created before this instant.
limit	number	No	Page size (default 50, max 200).
cursor	string	No	Opaque cursor from a previous response’s pagination.nextCursor.

Returns

Promise<ListDatasetsResponse> with shape { data: DatasetSummary[]; pagination: { limit, returned, hasMore, nextCursor } }.

Example

const { data, pagination } = await adaline.datasets.list({
  projectId: 'project_abc123',
  sort: 'createdAt:desc',
  limit: 50,
});

for (const dataset of data) {
  console.log(dataset.id, dataset.title);
}

if (pagination.hasMore) {
  const next = await adaline.datasets.list({
    projectId: 'project_abc123',
    cursor: pagination.nextCursor,
  });
}

---

create()

Create a new dataset. You can seed initial columns in the same call.

create(options: { dataset: CreateDatasetRequest }): Promise<Dataset>

Parameters

Name	Type	Required	Description
dataset	CreateDatasetRequest	Yes	Dataset definition.

CreateDatasetRequest shape (abbreviated):

interface CreateDatasetRequest {
  projectId: string;
  title: string;
  icon?: BaseEntityIcon;
  columns?: Array<{
    name: string;
    type: 'input' | 'output' | 'metadata' | /* dynamic types */;
    settings?: Record<string, unknown>;
  }>;
}

Returns

Promise<Dataset> — the created dataset with every column assigned a server-generated id.

Example

const dataset = await adaline.datasets.create({
  dataset: {
    projectId: 'project_abc123',
    title: 'Support triage eval set',
    icon: { type: 'emoji', value: '📚' },
    columns: [
      { name: 'question', type: 'input' },
      { name: 'expected', type: 'input' },
    ],
  },
});

console.log(`Created ${dataset.id} with ${dataset.columns.length} columns`);

---

get()

Fetch a single dataset (metadata + full column list, not rows).

get(options: { datasetId: string }): Promise<Dataset>

Parameters

Name	Type	Required	Description
datasetId	string	Yes	Dataset identifier.

Returns

Promise<Dataset> — see the Get dataset API schema.

Example

const dataset = await adaline.datasets.get({ datasetId: 'dataset_abc123' });

for (const column of dataset.columns) {
  console.log(column.name, column.type);
}

---

update()

Update dataset-level metadata (title, icon).

update(options: {
  datasetId: string;
  dataset: UpdateDatasetRequest;
}): Promise<DatasetSummary>

Parameters

Name	Type	Required	Description
datasetId	string	Yes	Dataset identifier.
dataset	UpdateDatasetRequest	Yes	Fields to update.

Returns

Promise<DatasetSummary> — the updated dataset metadata (note: returns the summary shape, not the full Dataset with columns).

Example

await adaline.datasets.update({
  datasetId: 'dataset_abc123',
  dataset: { title: 'Renamed dataset' },
});

---

delete()

Permanently delete a dataset and all of its columns and rows. Irreversible.

delete(options: { datasetId: string }): Promise<void>

Example

await adaline.datasets.delete({ datasetId: 'dataset_abc123' });

---

See Also

• DatasetRowsClient — adaline.datasets.rows.* row operations
• DatasetColumnsClient — adaline.datasets.columns.* column operations
• Adaline class — client construction
• PromptEvaluationsClient — run evaluations against a dataset
• API reference: List datasets · Create dataset · Get dataset · Update · Delete