{
  "id": "evaluator-optimizer",
  "code": "PTL-0069",
  "term": "Evaluator-Optimizer",
  "aliases": [],
  "category": "agents",
  "definition": "Evaluator-optimizer is a workflow loop in which one model call generates a response and another evaluates it against criteria and provides feedback, repeating until the output passes.",
  "description": "It works best when evaluation criteria are clear and when feedback demonstrably improves the output, as in literary translation or iterative search.",
  "example": null,
  "broader": [
    "agentic-workflow"
  ],
  "narrower": [],
  "related": [
    "self-refine",
    "llm-as-a-judge",
    "reflexion"
  ],
  "introduced": null,
  "sources": [
    {
      "title": "Building effective agents",
      "authors": "Anthropic",
      "year": 2024,
      "url": "https://www.anthropic.com/research/building-effective-agents"
    }
  ],
  "url": "https://protologue.com/t/evaluator-optimizer/",
  "citation": "Protologue. (2026). Evaluator-Optimizer. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.0.0, PTL-0069). https://protologue.com/t/evaluator-optimizer/"
}