POST
/
v4
/
guardrails
/
{id}
/
execute
curl --request POST \
  --url https://api.egp.scale.com/v4/guardrails/{id}/execute \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: <api-key>' \
  --data '{
  "messages": [
    {
      "role": "user",
      "content": "<string>"
    }
  ],
  "verbose": false
}'
{
  "is_safe": true,
  "score": 123,
  "severity": "low",
  "unsafe_categories": [
    {
      "category": "Cybersecurity & Hacking",
      "policy_id": "<string>",
      "threshold": 123,
      "is_safe": true,
      "description": "<string>",
      "score": 123,
      "severity": "low",
      "metadata": {}
    }
  ],
  "safe_categories": [
    {
      "category": "Cybersecurity & Hacking",
      "policy_id": "<string>",
      "threshold": 123,
      "is_safe": true,
      "description": "<string>",
      "score": 123,
      "severity": "low",
      "metadata": {}
    }
  ],
  "metadata": {},
  "verbose": false
}

Authorizations

​
x-api-key
string
header
required

Headers

​
x-selected-account-id
string | null

Path Parameters

​
id
string
required

Body

application/json

Execute a Guardrail to see if chat thread are compliant with configured GuardrailPolicy

Parameters: messages: List[Message] - list of messages to be checked verbose: bool - whether to include detailed information in the response, like safe categories and policy descriptions

​
messages
object[]
required
​
verbose
boolean
default:
false

Response

200
application/json
Successful Response
​
is_safe
boolean
required
​
score
number
required
​
unsafe_categories
object[]
required
​
safe_categories
object[]
required
​
severity
enum<string>
Available options:
low,
high
​
metadata
object
​
verbose
boolean
default:
false
List Attack Categories With Descriptions