# Keyless Agent Endpoints

The following endpoints are available for IDV Bridge On-Premise (Keyless Agent)

**1. `GET /health-check`**

**Purpose:** Check if the service is running.\
**Response:**

```json
{
  "version": "string",
  "status": "ok"
}
```

***

**2. `POST /v1/offline-enrollment`**

**Purpose:** Generate *client* and *server* enrollment states from an image. 

The Server state is required to register the new user on the Keyless backend, while Client state enables the [account recovery](https://docs.keyless.io/consumer/mobile-sdk-use-cases/guide-account-recovery) flow for ongoing authentication on a camera enabled device via the Mobile SDK.

\
**Request:**

* **Headers:**
  * `Exif-Transpose` *(boolean, optional)* — whether to rotate the image according to EXIF metadata.
  * `Content-Encoding` *(optional, const="base64")* — if the image is base64 encoded.
  * `Scenario` *(required)* — `"SELFIE" | "TRUSTED_SOURCE" | "DOCUMENT"`.
* **Body:** `image/*` — binary (JPEG, PNG, BMP, etc.)

**Responses:**

* `200`: Returns **OfflineEnrollment**
* `422`: Returns **FilterTriggered**
* `429`: Too many concurrent enrollments

`scenario` definitions are available [here](https://docs.keyless.io/idv-bridge/readme/idv-bridge-saas#scenarios).

***

**3. `POST /v1/online-enrollment`**

**Purpose:** Enroll a user directly from an image (without server state).

Client state enables the [account recovery](https://docs.keyless.io/consumer/mobile-sdk-use-cases/guide-account-recovery) flow for ongoing authentication on a camera enabled device via the Mobile SDK.

\
**Request:** Same headers/body as offline enrollment.\
**Responses:**

* `200`: Returns **OnlineEnrollment**
* `409`: Not configured for online enrollment
* `422`: Recognition filter error
* `429`: Too many concurrent enrollments

***

### Schemas

`OfflineEnrollment`

<table><thead><tr><th width="171.69549560546875">Field</th><th width="130.1904296875">Type</th><th>Description</th></tr></thead><tbody><tr><td><code>keylessId</code></td><td>string</td><td>Unique user ID</td></tr><tr><td><code>clientState</code></td><td>string</td><td>Opaque data for the Keyless SDK to allow a user to authenticate on a camera enabled device for the accoutn recovery flow.</td></tr><tr><td><code>serverState</code></td><td>string</td><td>Uncommitted server data; required to register the new user on the Keyless backend. </td></tr></tbody></table>

`OnlineEnrollment`

<table><thead><tr><th width="172.57537841796875">Field</th><th width="126.6259765625">Type</th><th>Description</th></tr></thead><tbody><tr><td><code>keylessId</code></td><td>string</td><td>Unique user ID</td></tr><tr><td><code>clientState</code></td><td>string</td><td>Opaque data for the Keyless SDK to allow a user to authenticate on a camera enabled device for the accoutn recovery flow.</td></tr></tbody></table>

`FilterTriggered`&#x20;

<table><thead><tr><th width="164.327392578125">Field</th><th width="139.609130859375">Type</th><th>Description</th></tr></thead><tbody><tr><td><code>title</code></td><td>string</td><td>Always <code>"Filter Triggered"</code></td></tr><tr><td><code>statusCode</code></td><td>integer</td><td>Always <code>422</code></td></tr><tr><td><code>detail</code></td><td>string</td><td>Explanation of why the filters failed</td></tr><tr><td><code>filters</code></td><td>object</td><td>See below</td></tr></tbody></table>

**`Filters`**

Each property represents a possible filter and has a status:

* `TRIGGERED` – filter was triggered (failure)
* `NOT_TRIGGERED` – filter passed
* `TURNED_OFF` – filter not active
* `CANNOT_RUN_ON_FRAME` – preprocessing failure

**Filter fields:**

<table><thead><tr><th width="212.60906982421875">Filters</th><th>Description</th></tr></thead><tbody><tr><td><code>faceIsOccluded</code></td><td>The facial features are in part covered by something to the extent that a biometric template could not be detected.</td></tr><tr><td><code>faceMissing</code></td><td>No face was detected in the image.</td></tr><tr><td><code>faceMultiple</code></td><td>More than one face was detected.</td></tr><tr><td><code>facePartial</code></td><td>Only a partial face was detected, for example the selfie may have been off-center.</td></tr><tr><td><code>faceTooSmall</code></td><td>A face was detected, but the face was too small to allow an accurate biometric template to be generated.</td></tr><tr><td><code>imageBlackAndWhite</code></td><td>The image is black and white.</td></tr></tbody></table>

#### Example 422 Error

```json
{
  "title": "Filter Triggered",
  "statusCode": 422,
  "detail": "Face occluded by object",
  "filters": {
    "faceIsOccluded": "TRIGGERED",
    "faceMissing": "NOT_TRIGGERED",
    "faceMultiple": "NOT_TRIGGERED",
    "facePartial": "NOT_TRIGGERED",
    "faceTooSmall": "NOT_TRIGGERED",
    "imageBlackAndWhite": "NOT_TRIGGERED"
  }
}

```