Moderate: user post safety verdict

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.

// Moderate: user post safety verdict — Jev judge eval (typesafe-ai/jev).
// A heated but non-violent forum post; the judge issues a safety verdict.
// 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: "Post: You people defending this update are clueless. I have used this app for 6 years and you just deleted the one feature that made it usable. Revert it or I am gone — and I am taking my whole team with me.",
  questions: {
    "verdict": {
      "type": "choice",
      "instructions": "Give the moderation verdict for this post.",
      "criteria": {
        "safe": "No policy issue — criticism, complaints, and venting are allowed.",
        "borderline": "Rude or heated but not removable; watch the thread.",
        "unsafe": "Hate, threats, harassment of a person, or disallowed content — remove."
      }
    }
  },
});

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