{
  "id": 18399530,
  "title": "Why the model shouldn't be the one that says \"done\"",
  "url": "https://urgent.news/2026/11/05/why-the-model-shouldnt-be-the-one-that-says-done",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-11-05T23:36:37.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/bitphill/why-the-model-shouldnt-be-the-one-that-says-done-ibh"
  },
  "original_language": "en",
  "account": "The model should not be the one to declare work complete, according to the author of Loopsmith, an open-source project built with AI assistance. The model's self-certification of completion is unreliable, as it is the party with the most incentive to reach a positive conclusion. A better approach is for a separate checker, written in deterministic Rust code, to verify the completion of a task. Loopsmith takes a job description in a plain text file, specifying the desired outcome, verification method, and possible termination conditions. The checker, compiled separately from the model's description, becomes the gate that determines when the job is truly finished. The gate can also invalidate results if needed. Loopsmith ensures that the model's work ends when the gate confirms completion, preventing issues like missing sections or edited tests. The tool is MIT-licensed, with a 20-second demo available to showcase its functionality.",
  "summary": "Disclosure: this article was written with AI assistance and reviewed and edited by me, the tool's author. Loopsmith is my own open-source project. I have a few jobs I redo every week and care about getting right: a competitor roundup, a lead list, a research brief. Handing them to an AI agent works up to a point. The part that kept failing for me was the end of the job, not the work itself. The…",
  "key_points": [
    "The model should not declare work complete, says Loopsmith author.",
    "Separate checker in deterministic Rust verifies task completion.",
    "Loopsmith prevents issues by ending model's work upon gate confirmation."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}