Verify content

Verify content API endpoint allows to verify LLM response against your custom policies

POST /v1/{team_id}/verify

cURL

cURL

curl --request POST \
  --url https://api.naviml.com/v1/{team_id}/verify \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "messages": [
    {
      "content": "<string>"
    }
  ],
  "verifiers": [
    {
      "id": "<string>",
      "type": "policy-verifier",
      "parameters": {
        "threshold": 0.5
      }
    }
  ],
  "explain": true
}
'

200

{
  "request_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "result": {
    "verifiers": [
      {
        "id": "<string>",
        "result": "<string>",
        "explanation": {
          "text": "<string>"
        },
        "details": {},
        "usage": {
          "prompt_tokens": 123,
          "completion_tokens": 123,
          "total_tokens": 123
        }
      }
    ]
  }
}

This endpoint allows to run multiple verifiers against a single LLM response. The response will contain the result of each verifier and the final action based on the verifiers’ results.

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

team_id · string · required

Team ID of the team owning the resources and under which the verifiers are executed

Body

application/json

messages · object[] · required

List of messages to run verifiers against

Show child attributes

verifiers · object[] · required

List of verifiers to execute with configurations

Option 1

Option 2

Show child attributes

explain · boolean

Includes explanation for the result in response

Response

200

*/*

OK

request_id · string<uuid> · required

Unique identifier for the request

action · enum<string> · required

Suggested action based on the result

Available options: allow, modify, deny

result · object · required

Show child attributes