A category schema is a structured description of the decision factors that matter when narrowing a particular product or service category.
It is not the Offer JSON Schema and not the category taxonomy. In AON's MCP and Skill workflows, a category schema can guide an AI product to ask focused follow-up questions before calling Offer Search.
Category schemas in AON MCP workflows
AON's category-schema pattern helps an MCP host or Skill understand what to ask before searching. The host remains responsible for the conversation, while AON's search contract remains the authority for accepted inputs and returned Offers.
See the Category Schema guideHow a category schema improves an Offer search
Resolve the category
The product identifies the supported AON category or uses the category-resolution flow instead of inventing an ID.
Read decision factors
The category schema returns the dimensions that commonly affect a useful decision in that category.
Ask focused questions
The AI product asks only the questions needed to clarify the user's current request and lets the user supply the answers.
Search with enriched intent
The product sends the clarified intent and supported constraints to the Offer search flow, preserving provenance and omitting unknown fields.
Implement Category schema safely
Keep
- The distinction between a category schema, the category taxonomy, and the Offer JSON Schema.
- The user's actual answers and their provenance when enriching the next Offer Query.
- Only supported category IDs, decision factors, and constraints from the active contract.
Do not
- Treating every option exposed by a schema as a user-selected constraint.
- Inventing missing preferences or using the schema as an eligibility, ranking, price, or checkout authority.
- Calling a category schema the Offer JSON Schema or the canonical category taxonomy.
Category schema, taxonomy, and Offer schema
These three concepts describe different layers of an AI offer workflow and should not be used interchangeably.
| Concept | Meaning |
|---|---|
| Category schema | Describes which decision factors are useful for clarifying a category-specific user request. |
| Category taxonomy | Defines the canonical hierarchy and IDs used to classify Offers and apply supported category constraints. |
| Offer schema | Defines the fields and shape of the commercial Offer returned to an Agent or supplied by a Partner. |
| Query guidance | Connects the factors to a follow-up interaction, but does not authorize the product to invent missing preferences or force a result. |
- A category schema is not a JSON schema for validating the Public Offer payload.
- A category schema does not decide eligibility, ranking, price, availability, or merchant checkout acceptance.
- Decision factors are guidance for clarification; the user's actual answer takes priority over a default or inferred value.
Category schema vs category taxonomy
A category taxonomy answers how offers are classified and which canonical IDs exist. A category schema answers which questions or decision factors may help an AI product understand a user's request within one category. See the Category taxonomy definition.
| Dimension | Category schema | Category taxonomy |
|---|---|---|
| Purpose | Supports request clarification and intent enrichment. | Supports Offer classification and deterministic category filtering. |
| Output | Decision factors and possible options for one category. | Canonical category IDs and hierarchy. |
| When it is used | Before search when the user's request needs clarification. | When classifying an Offer or applying an explicit category constraint. |
Example in practice
A user asks for 'some headphones.' The AI product reads the electronics category schema, asks whether the user wants over-ear or earbuds, their use case, budget, and important features, then searches with the answers the user actually provided. It does not treat every suggested option as a user requirement.
Common questions about category schema
Does a category schema replace the Offer Query?
No. A category schema is a guidance step that can make the next Offer Query more precise. The Query contract still defines the accepted request and response fields.
Is a category schema the same as a product taxonomy?
No. A taxonomy classifies products or Offers. A category schema describes decision factors that help an AI product clarify the user's request.
Can an AI product fill in every decision factor from an AON category schema automatically?
No. An AI product should preserve what the user stated, label bounded inferences correctly, and omit unknown values rather than inventing preferences.
Is a category schema the same as the Offer JSON Schema?
No. A category schema describes decision factors that help an AI product clarify a request before search. The Offer JSON Schema defines the fields and structure of a commercial Offer payload.
Sources
- AON MCP toolsAON documentation
- AON Category TaxonomyAON documentation
- JSON Schema specificationPrimary standard