term:agentic-loop

{
  "id": "term:agentic-loop",
  "_kind": "Term",
  "_file": "terminology/terms/terms-articles-and-refs.yaml",
  "_cluster": "terminology",
  "attributes": {
    "canonicalDefinitionId": "# TODO: link canonical definition",
    "displayName": "Agentic Loop",
    "kind": "concept",
    "description": "The \"Loop 2\" of a Hybrid Agentic System: a tool-using LLM-powered\nreasoning loop that iterates until a local objective is reached\n(read objective + constraints, decide what evidence is needed,\ncall tools, update plan, produce output). Source:\ndocs/articles/Hybrid Agentic Systems_ Two-loops Control Plane Guide.md.\n"
  },
  "outgoingEdges": [
    {
      "from": "term:agentic-loop",
      "to": "term:orchestration-loop",
      "kind": "synonym_of",
      "attributes": {}
    }
  ],
  "incomingEdges": [
    {
      "from": "term:agent-core-impl",
      "to": "term:agentic-loop",
      "kind": "subsumes",
      "attributes": {}
    }
  ]
}