[
{
"id": "2e19b8bf-2a4f-4f3c-9b7e-0f10c9e55c01",
"name": "This is a domain layered system.",
"tags": ["intro", "layers", "models"],
"details": "This system is a series of features that are organized into cohesive domains. Each domain can also export CRUDS access to models, if it chooses to. Domains can be listed, features can be listed, described and called, and so can models."
},
{
"id": "7c0c9643-92f4-4603-984d-3a25b9bd3a28",
"name": "MCP navigation workflow (high level)",
"tags": ["mcp", "workflow", "discovery"],
"details": "Typical AI-driven navigation flow: (1) Call list_domains to discover business domains. (2) For the chosen domain, call list_features and list_models to understand available actions and data. (3) Optionally call describe_feature or describe_model to get precise input/output schemas. (4) Execute a feature with execute_feature, or operate on models with save/retrieve/delete/search/bulkInsert/bulkDelete. Prefer describe_* before execution to align arguments correctly."
},
{
"id": "0f9f5b67-9b8e-4b2a-8b28-2a8800d2c3a1",
"name": "Cross Layer Props (tracing across calls)",
"tags": ["mcp", "tracing", "crossLayerProps"],
"details": "crossLayerProps is an optional argument you can send with NIL MCP tool calls to enable end-to-end tracing across layers (features/services) and across multiple tool invocations. It carries correlation ids that the system logs at each hop so you can stitch together a full execution story.",
"example": {
"strategy": "Generate a single aiSessionId (uuid) for the entire AI session, and an aiRequestId (uuid) per user request. Reuse both ids for every tool call made while fulfilling that one user request.",
"idsObject": {
"crossLayerProps": {
"logging": {
"ids": [
{ "aiSessionId": "8c2ab2c0-6f8e-4a6b-9b4e-3e9c1e3a9d10" },
{ "aiRequestId": "f4c7a8e2-1b9c-49bf-9f07-6b9d7f4e2a33" }
]
}
}
},
"calls": [
{
"tool": "list_domains",
"input": {
"crossLayerProps": {
"logging": {
"ids": [
{ "aiSessionId": "8c2ab2c0-6f8e-4a6b-9b4e-3e9c1e3a9d10" },
{ "aiRequestId": "f4c7a8e2-1b9c-49bf-9f07-6b9d7f4e2a33" }
]
}
}
}
},
{
"tool": "list_features",
"input": {
"domain": "inventory",
"crossLayerProps": {
"logging": {
"ids": [
{ "aiSessionId": "8c2ab2c0-6f8e-4a6b-9b4e-3e9c1e3a9d10" },
{ "aiRequestId": "f4c7a8e2-1b9c-49bf-9f07-6b9d7f4e2a33" }
]
}
}
}
},
{
"tool": "execute_feature",
"input": {
"domain": "inventory",
"featureName": "reserveItem",
"args": { "sku": "ABC-123", "quantity": 1, "userId": "u-42" },
"crossLayerProps": {
"logging": {
"ids": [
{ "aiSessionId": "8c2ab2c0-6f8e-4a6b-9b4e-3e9c1e3a9d10" },
{ "aiRequestId": "f4c7a8e2-1b9c-49bf-9f07-6b9d7f4e2a33" }
]
}
}
}
}
],
"notes": "Do not create a new id per individual tool call during the same user request; that breaks correlation. Keep aiSessionId stable for the chat/session, rotate aiRequestId per top-level user instruction."
}
},
{
"id": "a0a1c3cf-7f3f-4e4b-9f1e-2b1a9a7b11aa",
"name": "list_domains",
"tags": ["mcp", "nil", "discovery"],
"details": "Lists domains shows the domains available for access.",
"example": {
"request": {
"tool": "list_domains",
"input": {}
},
"response": {
"domains": [
{ "name": "auth", "description": "User accounts and authentication" },
{ "name": "inventory", "description": "Stock and availability" }
]
},
"notes": "Start here. Use a domain's description to decide next actions."
}
},
{
"id": "f3b7b0d1-0d4e-4a25-9c76-8a7ef0f2d3c2",
"name": "list_features",
"tags": ["mcp", "nil", "discovery", "features"],
"details": "Lists callable business features for a domain. Use before execute_feature to get the exact input schema for a feature.",
"example": {
"request": {
"tool": "list_features",
"input": { "domain": "inventory" }
},
"response": {
"features": [
{
"name": "checkItemAvailability",
"description": "Check stock for an item"
},
{ "name": "reserveItem", "description": "Place a hold on an item" }
]
},
"notes": "Only function-typed exports are listed; hidden features are omitted."
}
},
{
"id": "3f2f4a9a-4d75-41f0-b1a9-3b7f5a223c17",
"name": "describe_feature",
"tags": ["mcp", "nil", "schema", "features"],
"details": "Returns an OpenAPI-like schema for a feature.",
"example": {
"request": {
"tool": "describe_feature",
"input": { "domain": "inventory", "featureName": "reserveItem" }
},
"response": {
"schema": {
"description": "Reserve an item for a user",
"properties": { "args": { "type": "object" } }
}
},
"notes": "Use before execute_feature to validate arguments."
}
},
{
"id": "9a1a7b5d-1b2e-4b6c-9f0c-2d3e4f5a6b7c",
"name": "execute_feature",
"tags": ["mcp", "nil", "execution", "features"],
"details": "Executes a specific domain feature. Pass feature-specific args and optional crossLayerProps for tracing.",
"example": {
"request": {
"tool": "execute_feature",
"input": {
"domain": "inventory",
"featureName": "reserveItem",
"args": { "sku": "ABC-123", "quantity": 1, "userId": "u-42" },
"crossLayerProps": {
"logging": { "ids": [{ "traceId": "123e4567" }] }
}
}
},
"response": {
"result": { "reservationId": "r-1001", "status": "RESERVED" }
},
"notes": "Errors are returned as structured error objects by the server."
}
},
{
"id": "b0c5c3f3-6a7d-4f0a-8b1c-6d3a2f1e9c88",
"name": "list_models",
"tags": ["mcp", "models", "discovery"],
"details": "Lists ORM-backed models for a domain with optional descriptions. Hidden models/domains are omitted.",
"example": {
"request": { "tool": "list_models", "input": { "domain": "inventory" } },
"response": {
"models": [
{
"modelType": "inventory/Inventories",
"description": "Inventory records"
}
]
},
"notes": "Model type strings encode namespace and plural name."
}
},
{
"id": "d7f2b1c0-1b9d-4ef3-88a9-1c2e3f4a5b6c",
"name": "describe_model",
"tags": ["mcp", "models", "schema"],
"details": "Returns the model schema for a given modelType. Use to learn required fields and types.",
"example": {
"request": {
"tool": "describe_model",
"input": { "domain": "inventory", "modelType": "inventory/Inventories" }
},
"response": {
"schema": {
"properties": {
"id": { "type": "string" },
"sku": { "type": "string" },
"quantity": { "type": "number" }
}
}
},
"notes": "modelType is parsed into namespace and plural model name."
}
},
{
"id": "c5b7e0a2-4a6f-4a4e-8629-0f1a2b3c4d5e",
"name": "Model CRUD: save",
"tags": ["mcp", "models", "crud", "save"],
"details": "Creates or updates a model instance. Pass modelType and instance data.",
"example": {
"request": {
"tool": "save",
"input": {
"modelType": "inventory/Inventories",
"instance": { "sku": "ABC-123", "quantity": 10 }
}
},
"response": { "id": "inv-1", "sku": "ABC-123", "quantity": 10 },
"notes": "Validation errors are returned as structured error objects."
}
},
{
"id": "e1d2c3b4-a5f6-4a7b-8c9d-0e1f2a3b4c5d",
"name": "Model CRUD: retrieve",
"tags": ["mcp", "models", "crud", "retrieve"],
"details": "Retrieves a single instance by id.",
"example": {
"request": {
"tool": "retrieve",
"input": { "modelType": "inventory/Inventories", "id": "inv-1" }
},
"response": { "id": "inv-1", "sku": "ABC-123", "quantity": 10 },
"notes": "If not found, server returns a model-not-found error object."
}
},
{
"id": "f6a5b4c3-d2e1-4c0b-9a8f-7e6d5c4b3a21",
"name": "Model CRUD: delete",
"tags": ["mcp", "models", "crud", "delete"],
"details": "Deletes an instance by id.",
"example": {
"request": {
"tool": "delete",
"input": { "modelType": "inventory/Inventories", "id": "inv-1" }
},
"response": null,
"notes": "Returns null on success."
}
},
{
"id": "1a2b3c4d-5e6f-4701-8a9b-0c1d2e3f4a5b",
"name": "Model CRUD: search",
"tags": ["mcp", "models", "crud", "search"],
"details": "Searches model instances via ORM query format.",
"example": {
"request": {
"tool": "search",
"input": {
"modelType": "inventory/Inventories",
"query": {
"filters": [{ "field": "sku", "op": "=", "value": "ABC-123" }]
}
}
},
"response": {
"instances": [{ "id": "inv-1", "sku": "ABC-123", "quantity": 10 }],
"page": { "size": 25, "number": 1, "total": 1 }
},
"notes": "The exact search format depends on the ORM configured."
}
},
{
"id": "6b7c8d9e-0f1a-42b3-8c4d-5e6f7a8b9c0d",
"name": "Model CRUD: bulkInsert",
"tags": ["mcp", "models", "crud", "bulk"],
"details": "Inserts many instances in a single call.",
"example": {
"request": {
"tool": "bulkInsert",
"input": {
"modelType": "inventory/Inventories",
"items": [
{ "sku": "ABC-123", "quantity": 5 },
{ "sku": "XYZ-999", "quantity": 2 }
]
}
},
"response": null,
"notes": "Returns null on success; validate inputs with describe_model first."
}
},
{
"id": "7c8d9e0f-1a2b-43c4-9d5e-6f7a8b9c0d1e",
"name": "Model CRUD: bulkDelete",
"tags": ["mcp", "models", "crud", "bulk"],
"details": "Deletes many instances by ids.",
"example": {
"request": {
"tool": "bulkDelete",
"input": {
"modelType": "inventory/Inventories",
"ids": ["inv-1", "inv-2"]
}
},
"response": null,
"notes": "Use search first to decide which ids to delete."
}
},
{
"id": "0a1b2c3d-4e5f-4607-8a90-b1c2d3e4f506",
"name": "Putting it together: example conversation flow",
"tags": ["mcp", "workflow", "examples"],
"details": "Example of an AI deciding which tools to call and in what order.",
"example": {
"markdown": "User asks: \"Reserve 1 unit of ABC-123 if available.\"\n\n1. list_domains {}\n - Choose domain = inventory.\n2. list_features { domain: \"inventory\" } and list_models { domain: \"inventory\" }\n - See feature reserveItem and model Inventories.\n3. describe_feature { domain: \"inventory\", featureName: \"checkItemAvailability\" }\n - Confirm args: { sku: string, quantity: number }.\n4. execute_feature { domain: \"inventory\", featureName: \"checkItemAvailability\", args: { sku: \"ABC-123\", quantity: 1 }, crossLayerProps: { logging: { ids: [{ aiSessionId: \"8c2ab...\" }, { aiRequestId: \"f4c7a...\" }] } } }\n - If available: true -> proceed.\n5. execute_feature { domain: \"inventory\", featureName: \"reserveItem\", args: { sku: \"ABC-123\", quantity: 1, userId: \"u-42\" }, crossLayerProps: { logging: { ids: [{ aiSessionId: \"8c2ab...\" }, { aiRequestId: \"f4c7a...\" }] } } }\n - Returns reservationId.\n"
}
},
{
"id": "0a1b2c3d-4e5f-4607-8a90-b1c2d3e4f507",
"name": "Working with the user",
"tags": ["user", "interaction"],
"details": "Unless the user asks for it don't explain Domain/Models of the system to the user. The features are the primary thing the user cares about, and only to the extent that YOU the AI call them. The details about what goes in them, only matters for them to run other features first. Knowing how to call and navigate this interface is YOUR (the AI)'s job."
}
]
