{
  "handle": "MetaPrompt",
  "mechanism": "A higher-order {{meta}} {{prompt}}ing technique where prompts are used to generate, refine, or analyze other prompts rather than directly answering questions. The LLM acts as a prompt engineer, creating task-specific prompts from templates or improving existing prompts through critique. Enables reusable prompt templates that generalize across problem categories. It constructs a {{prompt_chain}} where the output of the first stage is the prompt instruction for the second.",
  "gloss": "Use prompts to generate or refine other prompts",
  "failure_modes": [
    "{{prompt}} Drift: Generated prompts diverge from intent over iterations.",
    "{{meta}}-Complexity: {{meta}}-prompt harder to write than direct prompt.",
    "Overfitting: Generated prompts too specific to examples."
  ],
  "invariants": [
    "Output is a prompt",
    "{{prompt}} instructs how to prompt"
  ],
  "preconditions": [
    "{{prompt}} engineering principles",
    "{{task}} description"
  ],
  "postconditions": [
    "Optimized prompt generated"
  ],
  "_meta": {
    "tier": 1,
    "ring": 2,
    "supersedes": [
      "sema:MetaPrompt#mh:SHA-256:7a3a4dd5ba4513c3636d3d018e632f7503cc77e0d0bbb946975e00ba8257be92",
      "sema:MetaPrompt#mh:SHA-256:a60668b9460c39064cec2e6b4a87e6f4d92e4164927ba453f3cf4ecdb7622f43",
      "sema:MetaPrompt#mh:SHA-256:a66501101ec29854bb461a93e28bea471b5726a17085db26987c60ac65c1488d",
      "sema:MetaPrompt#mh:SHA-256:db51ab77fd1592155749c401b495f1090c0bc820458fe34b2412ce040c7e70ef",
      "sema:MetaPrompt#mh:SHA-256:f99473d1cac3cde3e39a0428c73a91c366925a3845747631a8f89a01137b582d",
      "sema:MetaPrompt#mh:SHA-256:308ceb209aa848e10d8d3693adf4ad8aea7d957d918958034acc1a78524e8fcc"
    ],
    "path": [
      "Mind",
      "Reasoning"
    ]
  },
  "sema_id": "sema:MetaPrompt#mh:SHA-256:5d117201e4b2fe776f32e4effb8bb9b0d541e361ba37468250283f277026c974",
  "sema_ref": "MetaPrompt#5d11",
  "sema_stub": "5d11",
  "signature": [
    "Meta(Prompt)"
  ],
  "dependencies": {
    "references": {
      "meta": "sema:Meta#mh:SHA-256:f277723b905f6c94549f27f7eb58597096f3052165f5ad90af8b273188cee079",
      "prompt": "sema:Prompt#mh:SHA-256:1b04d34a0c55d6d2c6fea4f1df4f517073ede2f20f1f8d1d149419010b440a0e",
      "prompt_chain": "sema:PromptChain#mh:SHA-256:02b5f2277c89d51e35336a8735f5a8bf26304332db887961bb23790c8c82b2f9",
      "task": "sema:Task#mh:SHA-256:a85010f508f85055170c396c8fa390125c1206888c565ac10651f3969a339f8d"
    }
  },
  "sema_layer": "Mind",
  "sema_category": "Reasoning",
  "used_by": {}
}