Skip to main content

Send a Synchronous Scan Request

POST 

/v1/scan/sync/request

Post a scan request containing prompt/model-response that returns a synchronous scan response

Request

Header Parameters

    x-pan-token stringrequired

    API key token

Body

required

Scan request object

    tr_id string

    Unique identifier for the transaction correlating prompt and response

    ai_profile objectrequired

    AI profile to use for the scan. You can specify one of the following - profile_id or profile_name

    profile_id string

    Unique identifier for the profile. If not provided, then profile_name is required.

    profile_name string

    Name of the profile. If not provided, then profile_id is required.

    metadata object

    Optionally send the app_name, app_user, and ai_model in the metadata

    app_name string

    AI application requesting the content scan

    app_user string

    End user using the AI application

    ai_model string

    AI model serving the AI application

    contents object[]required

    List of prompt or response or prompt/response pairs. The last element is the one that needs to be scanned, and the previous elements are the context for the scan.

  • Array [
  • prompt string

    The prompt content that you want to scan

    response string

    The response content that you want to scan

  • ]

Responses

successfully scanned request

Schema
    report_id stringrequired

    Unique identifier for the scan report

    scan_id uuidrequired

    Unique identifier for the scan

    tr_id string

    Unique identifier for the transaction

    profile_id uuid

    Unique identifier of the AI security profile used for scanning

    profile_name string

    AI security profile name used for scanning

    category stringrequired

    Category of the scanned content verdicts such as "malicious" or "benign"

    action stringrequired

    The action is set to "block" or "allow" based on AI security profile used for scanning

    prompt_detected object
    url_cats boolean

    Indicates whether prompt contains any malicious URLs

    dlp boolean

    Indicates whether prompt contains any sensitive information

    injection boolean

    Indicates whether prompt contains any injection threats

    response_detected object
    url_cats boolean

    Indicates whether response contains any malicious URLs

    dlp boolean

    Indicates whether response contains any sensitive information

    created_at date-time

    Scan request timestamp

    completed_at date-time

    Scan completion timestamp

Loading...