Classify: collections email tone

The job

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

The prompt

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

// Classify: collections email tone — Jev judge eval (typesafe-ai/jev).
// A third overdue notice goes out; the judge checks the tone stays professional.
// 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: "Subject: Third notice — invoice INV-2210 is 45 days overdue\nBody: Our records show $4,120 outstanding for 45 days. Please remit within 7 days to avoid service suspension. If payment is already sent, reply with the reference and we will reconcile immediately.",
  questions: {
    "tone": {
      "type": "choice",
      "instructions": "Classify the tone of this collections email.",
      "criteria": {
        "professional": "Firm but courteous; states facts, deadline, and a remedy.",
        "passive": "Vague or apologetic to the point of hiding the deadline.",
        "hostile": "Threatening, insulting, or abusive language."
      }
    }
  },
});

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