Refund chat: agent stated the amount before issuing it

The job

Jev judge eval (1 boolean) over a fixed state : runnable via experimental_evaluate.

The prompt

Copy this into Muse and replace anything in [square brackets] with your own details.

// Refund chat: agent stated the amount before issuing it — Jev judge eval (typesafe-ai/jev).
// A support agent issues a $42.50 refund; the judge checks the exact amount was stated before the refund was confirmed.
// Model: typesafe-ai/jev via the Vercel AI Gateway ($0.042/1M input tokens).
// Auth: set AI_GATEWAY_API_KEY (vck_…) in the environment — never paste the key here.
// For zero-data-retention / no-training workloads, add
// providerOptions: { gateway: { … } } to the evaluate call below.

import { experimental_evaluate } from 'ai';

const result = await experimental_evaluate({
  model: 'typesafe-ai/jev',
  state: "Customer: My order arrived damaged, I want a refund.\nAgent: I am sorry about that. I can refund order #88121 in full: $42.50 back to your Visa ending 4412 within 3-5 business days. Shall I go ahead?\nCustomer: Yes please.\nAgent: Done — $42.50 refunded to your Visa ending 4412. Confirmation R-99210.",
  questions: {
    "amount_disclosed": {
      "type": "boolean",
      "instructions": "True only if the agent stated the exact refund amount ($42.50) before the customer approved the refund."
    }
  },
});

console.log(JSON.stringify(result, null, 2));