Enable/disable a guardrail type for agent

PUT

/api/v2/agents/{agent_id}/guardrails

Enable or disable a specific guardrail type for an agent.
This will create a new guardrail instance if enabled and not already created.

Staging behaviour:

Enable or disable a specific guardrail type for the staging version instead of the published agent.

Changes are propagated to the published agent via the Publish API(POST api/v2/agents/{agent_id}/publish).

Path Parameters:
agent_id (int): ID of the agent

Request Body:
guardrail_type (str): Type of guardrail to enable/disable
enable (bool): Whether to enable or disable this guardrail type
config (dict, optional): Additional data for specific guardrail types

Returns:
JSON response with the enabled/disabled guardrail

Request

Path Params

Header Params

Body Params

application/json

Required

Examples

Responses

🟢200OK

application/json

Bodyapplication/json

🟠400Bad Request

🟠403Record Not Found

🟠404Server Error

🔴500

🔴503

Request Example

Shell

JavaScript

Java

Swift

curl --location --globoff --request PUT 'https:///api/v2/agents//guardrails' \
--header 'Organization-Id;' \
--header 'Ocp-Apim-Subscription-Key: ' \
--header 'Authorization: Bearer ' \
--header 'Content-Type: application/json' \
--data '{
    "guardrail_type": "topic_control",
    "enable": true
}'

Response Example

200 - Successful Response

{
    "success": true,
    "message": "Guardrail enabled successfully",
    "data": {
        "guardrail": {
            "id": 1,
            "name": "Content Boundaries",
            "description": "Restricts discussions to safe topics",
            "created_on": "2024-03-15T10:30:00Z",
            "modified_on": "2024-03-16T09:15:00Z",
            "created_by": "[email protected]",
            "modified_by": "[email protected]",
            "organization_id": 1,
            "guardrail_type": "topic_control",
            "is_enabled": true,
            "is_default": false,
            "is_deleted": false,
            "guidelines": [
                "No discussions about illegal activities",
                "No harmful content"
            ]
        }
    }
}

Apply Guardrails on Input Text

Get Guardrails