{
  "schema": "vedokrok.public-item.v1",
  "release_id": "MHC-RPUB-20260920-75ad787a",
  "url": "/knowledge/add-planner-builder-evaluator-roles-only-after-the-simple-loop-hits-a-ceiling",
  "id": "MHC-D-RESEARCH-0715",
  "version": "0.1.0",
  "title": "Add planner-builder-evaluator roles only after the simple loop hits a ceiling",
  "summary": "Three agents are not automatically one agent with better judgment.",
  "kind": "heuristic",
  "body": "Start with the simplest capable worker and a clear test. Add a planner when decomposition itself is a recurring failure, and add an evaluator when independent quality feedback improves results. Measure the new loop against the simpler baseline. Keep role boundaries narrow enough that handoffs carry artifacts and decisions rather than duplicated conversation.",
  "limits": [
    "Extra agents add latency, cost and new coordination failures; simplify again when a stronger model makes a role redundant."
  ],
  "topics": [
    "union-ai-coding-harness-work"
  ],
  "intents": [],
  "source_ids": [
    "RS-4CF3167EB387E6BE"
  ],
  "evidence": [
    {
      "claim": "Anthropic recommends increasing harness complexity only when simpler approaches fail, and used component removal to identify which harness pieces were load-bearing.",
      "source_id": "RS-4CF3167EB387E6BE",
      "role": "supports",
      "note": "Ablation can miss interactions between components; test representative end-to-end tasks.",
      "locator": "Simplification and methodical component removal"
    }
  ],
  "use_when": [
    "A single-agent workflow is failing on decomposition or quality and you are considering a multi-role architecture."
  ],
  "avoid_when": [
    "Extra agents add latency, cost and new coordination failures; simplify again when a stronger model makes a role redundant."
  ],
  "example": "A full-stack build may justify planner, builder and evaluator roles; a five-line configuration change probably does not.",
  "check": "Each added agent role exists because it fixes a measured failure mode, not because multi-agent architecture sounds advanced.",
  "sources": [
    {
      "id": "RS-4CF3167EB387E6BE",
      "title": "Harness design for long-running application development",
      "url": "https://www.anthropic.com/engineering/harness-design-long-running-apps"
    }
  ],
  "relations": [],
  "collections": [
    {
      "id": "RC-A76353E5E9A33FE5",
      "title": "Make coding agents work in inspectable increments",
      "url": "/collections/make-coding-agents-work-in-inspectable-increments"
    }
  ]
}