ObjectSchema extension type

A JSON Schema definition for an object with properties.

ObjectSchema is used to define the expected structure, data types, and constraints for MCP argument objects. It allows you to specify:

• Which properties an object can or must have (properties, required).
• The schema for each of those properties (e.g., string, number, nested object).
• Whether additional properties not explicitly defined are allowed (additionalProperties, unevaluatedProperties).
• Constraints on the number of properties (minProperties, maxProperties).
• Constraints on property names (propertyNames).

See https://json-schema.org/understanding-json-schema/reference/object.html for more details on object schemas.

Example:

To define a schema for a product object that requires productId and productName, has an optional price (non-negative number) and optional tags (list of unique strings), and optional dimensions (an object with required numeric length, width, and height):

final productSchema = ObjectSchema(
  title: 'Product',
  description: 'Schema for a product object',
  required: ['productId', 'productName'],
  properties: {
    'productId': Schema.string(
      description: 'Unique identifier for the product',
    ),
    'productName': Schema.string(description: 'Name of the product'),
    'price': Schema.num(
      description: 'Price of the product',
      minimum: 0,
    ),
    'tags': Schema.list(
      description: 'Optional list of tags for the product',
      items: Schema.string(),
      uniqueItems: true,
    ),
    'dimensions': ObjectSchema(
      description: 'Optional product dimensions',
      properties: {
        'length': Schema.num(),
        'width': Schema.num(),
        'height': Schema.num(),
      },
      required: ['length', 'width', 'height'],
    ),
  },
  additionalProperties: false, // No other properties allowed beyond those defined
);

This schema can then be used with the validate method to check if a given JSON-like map conforms to the defined structure.

For example, valid data might look like:

{
  "productId": "ABC12345",
  "productName": "Super Widget",
  "price": 19.99,
  "tags": ["electronics", "gadget"],
  "dimensions": {"length": 10, "width": 5, "height": 2.5}
}

And invalid data (e.g., missing productName, or an extra undefined property):

{
  "productId": "XYZ67890",
  "price": 9.99
}

{
  "productId": "DEF4567",
  "productName": "Another Gadget",
  "color": "blue" // Invalid if additionalProperties is false
}