{
  "name": "ChunkMaps",
  "alternateName": "CHUNKMAPS",
  "version": "1.0.0",
  "description": "Open taxonomy for ChunkMaps: hierarchical semantic mapping for human communication and agentic AI systems (domain, up/down/across vectors, four models, CHUNKS_OK).",
  "canonical_base": "https://chunkmaps.com/NLP_improved_communication/",
  "spec_urls": {
    "json": "https://chunkmaps.com/NLP_improved_communication/spec/chunkmaps.json",
    "yaml": "https://chunkmaps.com/NLP_improved_communication/spec/chunkmaps.yaml",
    "human": "https://chunkmaps.com/NLP_improved_communication/chunkmaps_spec.html"
  },
  "author": {
    "name": "Tony Nudd BA MSc",
    "url": "https://www.linkedin.com/in/businessrobotics"
  },
  "domain": {
    "id": "domain",
    "definition": "The topic boundary that keeps chunking relevant. All up, down, and across moves must stay inside the Domain.",
    "required_with": ["desired_outcome"],
    "examples": ["Team meeting culture", "Border Security", "Vendor RFP selection"],
    "rules": [
      "State Domain explicitly before mapping",
      "Prune off-topic chunks that drift outside Domain",
      "Re-state Domain when the map grows wide"
    ]
  },
  "desired_outcome": {
    "id": "desired_outcome",
    "definition": "The agreement or action the Journey of Persuasion (or agent workflow) must land on.",
    "examples": ["Approve a 90-day decision-memo pilot", "Recommend one cloud vendor with residual risk below threshold"]
  },
  "vectors": {
    "up": {
      "id": "chunk_up",
      "label": "Chunk Up",
      "definition": "Move to higher meaning: purpose, shared values, category, policy intent, or strategic frame.",
      "typical_questions": [
        "This chunk is an example of what?",
        "This belongs to which category?",
        "What is the shared purpose or intent?"
      ],
      "page": "chunking_up.html"
    },
    "down": {
      "id": "chunk_down",
      "label": "Chunk Down",
      "definition": "Move to specifics: evidence, examples, procedures, and actionable detail.",
      "typical_questions": [
        "An example of this chunk is?",
        "Specifically how?",
        "What evidence supports this?"
      ],
      "page": "chunking_down.html"
    },
    "across": {
      "id": "chunk_across",
      "label": "Chunk Across",
      "definition": "Move laterally to peer alternatives, related cases, and options at the same level of abstraction.",
      "typical_questions": [
        "Another example of this is?",
        "What peer alternatives exist?",
        "What related cases sit at this level?"
      ],
      "page": "chunking_across.html"
    }
  },
  "working_memory": {
    "rule": "7_plus_or_minus_2",
    "label": "7 plus or minus 2",
    "definition": "Keep top-level siblings within human working-memory limits (about 5 to 9 chunks) so maps stay cognitively manageable for people and context windows.",
    "page": "7-plus-minus-2.html"
  },
  "models": [
    {
      "id": "model_1_definition",
      "number": 1,
      "name": "Definition Model",
      "definition": "Places chunks in a logical hierarchy of the world: category (up), examples (down), and peer examples (across).",
      "use_when": "First pass to locate where a chunk fits in a definitive logical hierarchy.",
      "page": "the_4_chunk_map_models.html"
    },
    {
      "id": "model_2_interaction",
      "number": 2,
      "name": "Interaction Model",
      "definition": "Maps what a chunk does: intention and results (up), specific how (down), and peer interaction examples (across).",
      "use_when": "After Model 1, to understand interactions between chunks.",
      "page": "the_4_chunk_map_models.html"
    },
    {
      "id": "model_3_environment",
      "number": 3,
      "name": "Environment Model",
      "definition": "Maps the environment around a chunk: where it lives, constraints, and contextual conditions that shape meaning and action.",
      "use_when": "To surface context, constraints, and environmental forces around mapped chunks.",
      "page": "the_4_chunk_map_models.html"
    },
    {
      "id": "model_4_features",
      "number": 4,
      "name": "Features Model",
      "definition": "Maps distinctive features and attributes of a chunk to deepen creativity and differentiation.",
      "use_when": "To extract features that enable new insights after Models 1-3.",
      "page": "the_4_chunk_map_models.html"
    }
  ],
  "chunks_ok": [
    {
      "letter": "C",
      "name": "Current Situation",
      "prompt": "Current Situation your audience is in"
    },
    {
      "letter": "H",
      "name": "Headaches",
      "prompt": "Headaches caused by the current situation"
    },
    {
      "letter": "U",
      "name": "Utopia",
      "prompt": "Utopia wanted by your audience for the future"
    },
    {
      "letter": "N",
      "name": "Next Steps",
      "prompt": "Next Steps - what they should do immediately"
    },
    {
      "letter": "K",
      "name": "Knowledge",
      "prompt": "Knowledge - why you can help / why your solution / why they benefit"
    },
    {
      "letter": "S",
      "name": "Solution Overview",
      "prompt": "Solution Overview - how you solve the headaches"
    },
    {
      "letter": "O",
      "name": "Offer",
      "prompt": "Offer - compelling reason to accept now"
    },
    {
      "letter": "K_close",
      "name": "Knowledge (close)",
      "prompt": "Knowledge (close) - reinforce credibility and benefit"
    }
  ],
  "outline_form": {
    "description": "Canonical text outline for a Domain Chunk Map.",
    "template": "DOMAIN: <topic>\nDESIRED OUTCOME: <agreement / action>\n├── [UP] abstract / purpose chunks\n├── [ACROSS] sibling options / related cases\n└── [DOWN] evidence / examples / details"
  },
  "extensions": {
    "persuasion_flow": {
      "id": "persuasion_flow",
      "definition": "Procedure for human-to-human influence: Domain + outcome, Chunk Map, Journey of Persuasion path, counterargument map, CHUNKS_OK draft.",
      "skill": "skills/chunkmap-persuasion-flow.md",
      "guide": "chunkmap_persuasion_flow.html",
      "page": "the_journey_of_persuasion.html"
    },
    "agentic": {
      "blackboard": {
        "id": "agent_blackboard",
        "definition": "Shared semantic state schema plus write contracts so multi-agent swarms read and write against one mapped intent.",
        "skill": "skills/chunkmap-agent-blackboard.md",
        "guide": "agent_orchestrators_chunkmaps.html"
      },
      "dag_prompt_trees": {
        "id": "dag_prompt_trees",
        "definition": "Directed Acyclic Graph of bounded prompt nodes (mission, inputs, tools, write-back, exit criteria) with dependency edges and no cycles.",
        "guide": "agent_orchestrators_chunkmaps.html"
      },
      "policy_to_pipeline": {
        "id": "policy_to_pipeline",
        "definition": "Compile unstructured enterprise policy into versioned, citable logic chunks for automated evaluation pipelines.",
        "skill": "skills/chunkmap-policy-to-pipeline.md",
        "guide": "enterprise_architects_chunkmaps.html"
      },
      "spatial_reasoning": {
        "id": "spatial_chunk_maps",
        "definition": "Multi-dimensional chunk maps (up/down/across/time) that govern context windows and constrain drift across multi-step LLM reasoning loops.",
        "guide": "prompt_engineers_chunkmaps.html"
      }
    }
  },
  "keywords": [
    "AI prompt chaining frameworks",
    "cognitive load management tools",
    "structured argument mapping for LLMs",
    "enterprise decision-making frameworks",
    "AI prompt engineering",
    "cognitive framing",
    "semantic architecture",
    "multi-agent systems"
  ]
}