# Start a device authorization

`POST /v1/public/device/code`

URL: https://docs.spacerun.co/docs/platform/reference/device-authorization/createDeviceCode

- **Operation ID:** `createDeviceCode`
- **Tags:** Device authorization
- **Base URL:** `http://localhost:8787`

RFC 8628 style. The device calls this unauthenticated, shows `user_code` / `verification_uri` to the user, then polls `POST /v1/public/device/token` with `device_code`. The user approves at the portal (`POST /v1/me/device/approve`). Codes live 900 seconds. Limited to 3 requests per minute per client IP (429 `rate_limited`). An invalid body returns 400 `{ "error": "invalid_request" }`.

**Workflow:** [Integration guide](https://docs.spacerun.co/docs/platform/guides/device-authorization).

**Authentication:** none (public).

**Rate limit:** 3 requests per 60s per ip.

## Authentication

- **Auth:** none (public endpoint, no token required).

## Rate limit

3 requests per 60 seconds per ip. Exceeding it returns `429`.

## Request body

Required: yes.

Content type: `application/json`

Schema: `object`

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `user_data` | `object` | no | - |
| `client_name` | `string` | no | Max length: 200 |

**Example**

```json
{
  "client_name": "Example CLI"
}
```

## Responses

### 200

Device and user codes issued.

- Header `x-request-id`: Request identifier; quote it when reporting an error.
Content type: `application/json`

Schema: `object (DeviceCodeResponse)`

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `device_code` | `string` | yes | Secret polled by the device; 64 hex chars. Never shown to the user. |
| `user_code` | `string` | yes | Short code the user types at `verification_uri`, formatted `XXXX-XXXX`. |
| `verification_uri` | `string<uri>` | yes | - |
| `verification_uri_complete` | `string<uri>` | yes | `verification_uri` with the user code prefilled. |
| `expires_in` | `integer` | yes | Seconds until the code expires (900). |
| `interval` | `integer` | yes | Minimum seconds between token polls (5). |

## Errors

### 400

Malformed body (OAuth-style error, not the standard error body).

- Header `x-request-id`: Request identifier; quote it when reporting an error.
Content type: `application/json`

Schema: `object (DeviceOAuthError)`

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `error` | `string` | yes | RFC 8628 style error string.. One of: `"invalid_request"`, `"unsupported_grant_type"`, `"authorization_pending"`, `"slow_down"`, `"expired_token"`, `"access_denied"`, `"invalid_grant"`, `"server_error"` |
| `interval` | `integer` | no | Current polling interval in seconds; set for `authorization_pending` and `slow_down`. |

### 429

Rate limit exceeded (`rate_limited`).

- Header `x-request-id`: Request identifier; quote it when reporting an error.
Content type: `application/json`

Schema: `object (ErrorBody)`

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `error` | `string` | yes | Stable machine-readable error code, e.g. `forbidden`, `not_found`. |
| `details` | `any` | no | Optional structured context. For 403 it is `{ perm }`. |

**Error response**

```json
{
  "error": "rate_limited"
}
```

### 500

Unexpected server error (`internal`). No details are returned.

- Header `x-request-id`: Request identifier; quote it when reporting an error.
Content type: `application/json`

Schema: `object (ErrorBody)`, fields as described above.

**Error response**

```json
{
  "error": "internal"
}
```

### 503

 Maintenance can refuse this write before authentication (degraded blocks session start and checkout; lockdown blocks other public/user writes).

- Header `x-request-id`: Request identifier; quote it when reporting an error.
- Header `Retry-After`: Maintenance retry delay in seconds; present for panic_mode. Example: `30`.
Content type: `application/json`

Schema: `object (MaintenanceErrorBody)`

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `error` | `string` | yes | One of: `"panic_mode"` |
| `level` | `string` | yes | One of: `"degraded"`, `"lockdown"` |
| `reason` | `string \| null` | yes | - |

**Maintenance lockdown**

```json
{
  "error": "panic_mode",
  "level": "lockdown",
  "reason": "Scheduled maintenance"
}
```

## Examples

### curl

```bash
curl -X POST 'http://localhost:8787/v1/public/device/code' \
  -H 'Content-Type: application/json' \
  --data '{"client_name":"Example CLI"}'
```

### TypeScript

```ts
const response = await fetch("http://localhost:8787/v1/public/device/code", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "client_name": "Example CLI"
  }),
});
if (!response.ok) throw new Error(`HTTP ${response.status}: ${await response.text()}`);
const data = response.status === 204 || response.status === 205
  ? undefined
  : response.headers.get("content-type")?.includes("json")
    ? await response.json()
    : await response.text();
```