> **Building with AI coding agents?** Install the authstack plugin with one command. This equips your agent with accurate Scalekit implementation patterns.
>
> **Recommended** (the `-y` flags skip prompts, so the command runs without an interactive terminal):
> ```bash
> npx -y @scalekit-inc/cli setup -y
> ```
>
> Global:
> ```bash
> npm install -g @scalekit-inc/cli
> scalekit setup -y
> ```
>
> Supports Claude Code, Cursor, GitHub Copilot, Codex + skills for other Agent Skills-compatible agents in `~/.agents/skills`.
> Skills: integrate-agentkit, implement-saaskit, add-mcp-oauth, implement-sso, implement-scim.
> [Full setup guide](https://docs.scalekit.com/dev-kit/build-with-ai/)

---

# Connect AI agents to Scale AI

The Scale AI connector lets your AI agent act in each user's Scale AI account. Each user connects their own Scale AI API key once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 20 tools.

**Authentication:** API Key
**Categories:** AI, Developer Tools
**Tools:** 20: 8 read, 10 write, 2 destructive
**Users sign in with:** API key
**Built by:** Scalekit connector
**Try it:** [Playground in the Scalekit dashboard](https://app.scalekit.com/ws/signup?sk_intent=playground&provider=SCALEAI)

## Setup

1. ### Install the SDK

   **Node.js**

   ```bash
   npm install @scalekit-sdk/node dotenv
   ```

   **Python**

   ```bash
   pip install scalekit-sdk-python python-dotenv
   ```

2. ### Set your credentials

   Add your Scalekit credentials to your `.env` file. Find values in **[app.scalekit.com](https://app.scalekit.com)** > **Developers** > **API Credentials**.

   ```sh title=".env"
   SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
   SCALEKIT_CLIENT_ID=<your-client-id>
   SCALEKIT_CLIENT_SECRET=<your-client-secret>
   ```

3. ### Create the Scale AI connection

   In **AgentKit > Connections**, create a Scale AI connection. The name you give it is the `connection_name` your code passes. See [Configure connections](/agentkit/connections/).

## Tools

Pass the exact name to `execute_tool`, with an input like each tool's example. To ask for a tool that's missing, use the [request form](https://scalekitsupport.portal.usepylon.com/forms/request-a-connector-tool).

**Node.js**

```typescript
const result = await actions.executeTool({
  toolName: 'scaleai_audits_list',
  toolInput: {},
  connector: 'scaleai',
  identifier: 'user_123',
})
```

**Python**

```python
result = actions.execute_tool(
    tool_name="scaleai_audits_list",
    tool_input={},
    connection_name="scaleai",
    identifier="user_123",
)
```

`result.data` is the app's response as JSON, and `result.execution_id` (`executionId` in Node.js) is the ID of the call.

### `scaleai_audits_list`

List Fixless Audits · Read-only

Get the fixless audits of a task, or one audit by ID, with their feedback items (errors, flags, comments). Returns audit objects.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `id` (`string`, optional): Audit ID to get. Example: 6560a1b2c3d4e5f6a7b8c9d0.
- `task_id` (`string`, optional): Task ID to get audits for. Example: 576ba74eec471ff9b01557cc.

Example input: `{}`

### `scaleai_batch_get`

Get Batch · Read-only

Get one batch by name with its project, status and callback. Returns the batch object. Use scaleai_batch_status_get for task counts.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `batch_name` (`string`, required): Batch name, from scaleai_batches_list. Example: kitten_labeling_2020-07.

Example input: `{"batch_name":"<batch_name>"}`

### `scaleai_batch_status_get`

Get Batch Status · Read-only

Get a batch's status and the count of its tasks by status (pending, completed, error, canceled). Returns the counts.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `batch_name` (`string`, required): Batch name, from scaleai_batches_list. Example: kitten_labeling_2020-07.

Example input: `{"batch_name":"<batch_name>"}`

### `scaleai_batches_list`

List Batches · Read-only

List batches, newest first, optionally by project, status or creation time, with limit and offset paging. Returns batches with name, project, status and, with detailed, task counts.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `detailed` (`boolean`, optional): Set to true to include progress details. Example: true.
- `end_time` (`string`, optional): Only batches created at or before this ISO 8601 time. Example: 2026-01-31.
- `limit` (`integer`, optional): Page size, up to 100. Example: 100.
- `offset` (`integer`, optional): Number of batches to skip. Example: 0.
- `project` (`string`, optional): Only batches of this project name. Example: kitten_labeling.
- `start_time` (`string`, optional): Only batches created at or after this ISO 8601 time. Example: 2026-01-01.
- `status` (`string`, optional): Only batches in this status. Example: in_progress. One of: `staging`, `in_progress`, `completed`.

Example input: `{}`

### `scaleai_project_get`

Get Project · Read-only

Get one project by name with its task type, current parameters and parameter history. Returns the project object.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `project_name` (`string`, required): Project name, from scaleai_projects_list. Example: kitten_labeling.

Example input: `{"project_name":"<project_name>"}`

### `scaleai_projects_list`

List Projects · Read-only

List all projects in the account, optionally only archived or only active ones. Returns projects with name, task type, parameters and creation time. Use scaleai_project_get for one project.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `archived` (`boolean`, optional): Set to true for archived projects only, false for active projects only. Example: false.

Example input: `{}`

### `scaleai_task_get`

Get Task · Read-only

Get one task by ID with its status, parameters, metadata and, when completed, the response (annotations). Returns the task object.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.

Example input: `{"task_id":"<task_id>"}`

### `scaleai_tasks_list`

List Tasks · Read-only

List tasks, newest first, filtered by project, batch, type, status, review status, unique ID, tags or created/completed/updated time, with token paging. Returns tasks and next_token.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `batch` (`string`, optional): Only tasks of this batch name. Example: kitten_labeling_2026-01.
- `completed_after` (`string`, optional): Only tasks completed at or after this ISO 8601 time. Example: 2026-01-10.
- `completed_before` (`string`, optional): Only tasks completed at or before this ISO 8601 time. Example: 2026-01-20.
- `customer_review_status` (`string`, optional): Only tasks with this audit result; several comma-separated values allowed (accepted, fixed, commented, rejected). Example: accepted.
- `end_time` (`string`, optional): Only tasks created at or before this ISO 8601 time. Example: 2026-01-31.
- `include_attachment_url` (`boolean`, optional): Set to true to include presigned attachment URLs. Example: true.
- `limit` (`integer`, optional): Page size, 1 to 100. Defaults to 100. Example: 100.
- `next_token` (`string`, optional): next_token from the previous page. Example: eyJpZCI6IjU3NmJhNzRlIn0.
- `project` (`string`, optional): Only tasks of this project name. Example: kitten_labeling.
- `start_time` (`string`, optional): Only tasks created at or after this ISO 8601 time. Example: 2026-01-01.
- `status` (`string`, optional): Only tasks in this status. Example: completed. One of: `completed`, `pending`, `canceled`.
- `tags` (`string`, optional): Only tasks with this tag; several comma-separated values allowed. Example: pilot.
- `type` (`string`, optional): Only tasks of this type. Example: imageannotation. One of: `imageannotation`, `segmentannotation`, `videoannotation`, `videoplaybackannotation`, `textcollection`, `namedentityrecognition`, `lidarannotation`, `lidarsegmentation`, `lidarlinking`, `lidartopdown`, `sensorfusion`, `multistage`.
- `unique_id` (`string`, optional): Only tasks with this unique_id; several comma-separated values allowed. Example: kitten-01.jpg.
- `updated_after` (`string`, optional): Only tasks updated at or after this ISO 8601 time. Example: 2026-01-10.
- `updated_before` (`string`, optional): Only tasks updated at or before this ISO 8601 time. Example: 2026-01-20.

Example input: `{}`

### `scaleai_batch_create`

Create Batch · Write

Create a batch in a project to group tasks; Rapid batches can be calibration or self-label batches. Returns the new batch. Add tasks with scaleai_task_create and, for Rapid or Studio, finalize with scaleai_batch_finalize.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Batch name, unique in the account. Example: kitten_labeling_2026-01.
- `project` (`string`, required): Name of the project the batch belongs to. Example: kitten_labeling.
- `calibration_batch` (`boolean`, optional): Rapid only: set to true for a calibration batch. Example: false.
- `callback` (`string`, optional): URL or email address notified when tasks complete. Example: https://example.com/hooks/scale.
- `self_label_batch` (`boolean`, optional): Rapid only: set to true for a self-label batch. Example: false.

Example input: `{"name":"<name>","project":"<project>"}`

### `scaleai_batch_finalize`

Finalize Batch · Write

Finalize a Scale Rapid or Studio batch so its tasks start being worked on (and billed). Returns the batch. A no-op for other accounts.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `batch_name` (`string`, required): Batch name, from scaleai_batches_list. Example: kitten_labeling_2020-07.

Example input: `{"batch_name":"<batch_name>"}`

### `scaleai_batch_prioritize`

Prioritize Batch · Write

Set the priority of a batch's tasks that have not started, from 10 (lowest) to 30 (highest). Returns the result.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `batch_name` (`string`, required): Batch name, from scaleai_batches_list. Example: kitten_labeling_2020-07.
- `priority` (`integer`, required): Priority from 10 (lowest) to 30 (highest). Example: 20.

Example input: `{"batch_name":"<batch_name>","priority":20}`

### `scaleai_project_create`

Create Project · Write

Create a project for one task type, with default task parameters such as instructions. Returns the new project. Tasks and batches reference it by name.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `name` (`string`, required): Project name, unique in the account. Example: kitten_labeling.
- `type` (`string`, required): Task type of every task in the project. Example: imageannotation. One of: `imageannotation`, `segmentannotation`, `videoannotation`, `videoplaybackannotation`, `textcollection`, `namedentityrecognition`, `lidarannotation`, `lidarsegmentation`, `lidarlinking`, `lidartopdown`, `sensorfusion`, `multistage`.
- `consensus_attempts` (`integer`, optional): Studio consensus projects only: number of attempts. Example: 3.
- `params` (`object`, optional): Default task parameters, for example {"instruction": "Label every kitten"}; task requests override them. Example: {"instruction": "Please label the kittens"}
- `pipeline` (`string`, optional): Studio only: standard_task or consensus_task. Example: standard_task. One of: `standard_task`, `consensus_task`.
- `rapid` (`boolean`, optional): Set to true for a Scale Rapid project (self-serve accounts); enterprise customers leave it out. Example: true.
- `studio` (`boolean`, optional): Set to true for a Scale Studio project. Example: false.

Example input: `{"name":"<name>","type":"imageannotation"}`

### `scaleai_project_params_set`

Set Project Parameters · Write

Set new default task parameters on a project; future tasks inherit them unless the task request overrides them. With patch true the new values are merged into the current ones. Returns the updated project.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `project_name` (`string`, required): Project name, from scaleai_projects_list. Example: kitten_labeling.
- `instruction` (`string`, optional): Instructions added to every task's own instruction. Example: Label every kitten, including partly hidden ones.
- `params` (`object`, optional): Other default task parameters, for example geometries or attachment settings. Example: {"geometries": {"box": {"objects_to_annotate": ["kitten"]}}}
- `patch` (`boolean`, optional): Set to true to merge with the current parameters instead of replacing them. Defaults to false. Example: true.

Example input: `{"project_name":"<project_name>"}`

### `scaleai_task_create`

Create Task · Write

Create a labeling task of a given type (image, segmentation, video, text collection, named entity recognition, lidar, sensor fusion or multi-stage) with instructions, attachments and type-specific parameters. Returns the task with its ID. Live-key tasks are done by people and billed; test-key tasks are not.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `type` (`string`, required): Task type; it decides which type-specific parameters are needed. Example: imageannotation. One of: `imageannotation`, `segmentannotation`, `videoannotation`, `videoplaybackannotation`, `textcollection`, `namedentityrecognition`, `lidarannotation`, `lidarsegmentation`, `lidarlinking`, `lidartopdown`, `sensorfusion`, `multistage`.
- `attachment` (`string`, optional): URL of the image, video or file to label. Example: https://example.com/images/kitten-01.jpg.
- `attachments` (`array`, optional): Attachments for multi-attachment types such as textcollection, each an object. Example: [{"type": "website", "content": "https://example.com"}]
- `batch` (`string`, optional): Batch name; required for Scale Rapid projects. Example: kitten_labeling_2026-01.
- `callback_url` (`string`, optional): URL or email address notified when the task completes. Example: https://example.com/hooks/scale.
- `clear_unique_id_on_error` (`boolean`, optional): Set to true to free the unique_id if the task errors. Example: true.
- `instruction` (`string`, optional): Markdown instructions for the labelers. Example: Draw a box around every kitten.
- `metadata` (`object`, optional): Your own key/value data stored with the task. Example: {"source": "camera-3"}
- `params` (`object`, optional): Type-specific parameters merged into the request, for example geometries for imageannotation, fields for textcollection, text and labels for namedentityrecognition. Example: {"geometries": {"box": {"objects_to_annotate": ["kitten"], "min_height": 10, "min_width": 10}}}
- `project` (`string`, optional): Project name; not needed when batch is given. Example: kitten_labeling.
- `tags` (`array`, optional): Up to 5 tags. Example: ["pilot"]
- `unique_id` (`string`, optional): Your own ID for the task, unique across the account; a repeat returns 409. Example: kitten-01.jpg.

Example input: `{"type":"imageannotation"}`

### `scaleai_task_metadata_set`

Set Task Metadata · Write

Replace a task's metadata with the given key/value data. Returns the task.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `metadata` (`object`, required): Key/value data to store on the task. Example: {"source": "camera-3", "reviewed": true}
- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.

Example input: `{"metadata":{"source":"camera-3","reviewed":true},"task_id":"<task_id>"}`

### `scaleai_task_tags_add`

Add Task Tags · Write

Add tags to a task; tags it already has are ignored. Returns the task's tags.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `tags` (`array`, required): Tags to add; empty strings are not allowed. Example: ["pilot", "night"]
- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.

Example input: `{"tags":["pilot","night"],"task_id":"<task_id>"}`

### `scaleai_task_tags_set`

Set Task Tags · Write

Replace all of a task's tags with the given list. Returns the task's tags.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `tags` (`array`, required): Complete list of tags for the task. Example: ["pilot"]
- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.

Example input: `{"tags":["pilot"],"task_id":"<task_id>"}`

### `scaleai_task_unique_id_set`

Set Task Unique ID · Write

Set or change a task's unique_id. Returns the task.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.
- `unique_id` (`string`, required): New unique ID, unique across the account. Example: kitten-01.jpg.

Example input: `{"task_id":"<task_id>","unique_id":"<unique_id>"}`

### `scaleai_task_cancel`

Cancel Task · Destructive

Cancel a pending task, optionally freeing its unique_id for reuse. Returns the task. Completed tasks cannot be canceled.

Read only: no. Destructive: yes. Idempotent (safe to retry): yes.

Inputs:

- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.
- `clear_unique_id` (`boolean`, optional): Set to true to clear the task's unique_id so it can be reused. Example: true.

Example input: `{"task_id":"<task_id>"}`

### `scaleai_task_unique_id_delete`

Delete Task Unique ID · Destructive

Remove a task's unique_id so the value can be used again. Returns the task.

Read only: no. Destructive: yes. Idempotent (safe to retry): yes.

Inputs:

- `task_id` (`string`, required): Task ID, from scaleai_tasks_list or the create response. Example: 576ba74eec471ff9b01557cc.

Example input: `{"task_id":"<task_id>"}`

## Related

Other AI connectors ([all 216](/agentkit/connectors/?category=ai)):

- [Chorus](/agentkit/connectors/chorus/): Scalekit connector, Username and password, 9 tools
- [Gong](/agentkit/connectors/gong/): Scalekit connector, OAuth, 69 tools
- [Jiminny](/agentkit/connectors/jiminny/): Scalekit connector, Access token, 27 tools


---

## More Scalekit documentation

| Resource | What it contains | When to use it |
|----------|-----------------|----------------|
| [/llms.txt](/llms.txt) | Structured index with routing hints per product area | Start here — find which documentation set covers your topic before loading full content |
| [/llms-full.txt](/llms-full.txt) | Complete documentation for all Scalekit products in one file | Use when you need exhaustive context across multiple products or when the topic spans several areas |
| [sitemap-0.xml](https://docs.scalekit.com/sitemap-0.xml) | Full URL list of every documentation page | Use to discover specific page URLs you can fetch for targeted, page-level answers |
