Trust & safety

Session risk analysis

Classify a session as low, medium, or high risk based on the supplied behavioral signals. Explain which observed signals influenced the classification. Do not infer sensitive personal characteristics.

Customize this service

Input

A JSON object containing your task data.

How it becomes your API

Add your examples and define your output. Pronto builds a candidate and checks it against separate test examples. Your service becomes available after platform review.

Integration guide

Starting output contract

{
  "type": "object",
  "properties": {
    "risk": {
      "type": "string",
      "enum": [
        "low",
        "medium",
        "high"
      ]
    },
    "signals": {
      "type": "array",
      "items": {
        "type": "string"
      }
    },
    "review_required": {
      "type": "boolean"
    }
  },
  "required": [
    "risk",
    "signals",
    "review_required"
  ],
  "additionalProperties": false
}