Interface: GenerateOntologyRequest

@kortexya/reasoninglayer / GenerateOntologyRequest

Interface: GenerateOntologyRequest

Defined in: src/types/ontology.ts:50

Request body for ontology generation.

Remarks

Supports multi-turn clarification via session_id and answers. On first call, provide only prompt. If the response status is "needs_clarification", supply session_id and answers on the next call.

With stage set, the response is a "patch" for that stage alone.

Properties

answers?

optional answers?: object | null

Defined in: src/types/ontology.ts:52

Answers to clarification questions (2nd call).

draft?

optional draft?: object | null

Defined in: src/types/ontology.ts:82

Staged builder: the current draft scenario (including scope and use_cases) the patch extends.

Remarks

Sent verbatim: the draft is scenario JSON whose keys are the engine’s own snake_case names and the user’s feature names, so it is never re-cased.

language?

optional language?: string | null

Defined in: src/types/ontology.ts:64

The language of every human-readable text the generator writes: descriptions, labels, questions, notes and constraint messages.

Remarks

A BCP-47 tag or a language name ('en', 'fr-FR', 'English'). Sort and feature names stay English snake_case. Absent ⇒ the language of the prompt, which the model infers.

prompt

prompt: string

Defined in: src/types/ontology.ts:54

The user’s task description.

sessionId?

optional sessionId?: string | null

Defined in: src/types/ontology.ts:66

Session ID for multi-turn clarification.

stage?

optional stage?: OntologyStage | null

Defined in: src/types/ontology.ts:73

Staged builder: generate only this stage’s patch. Absent ⇒ one-shot generation of a whole scenario.

tenantId?

optional tenantId?: string | null

Defined in: src/types/ontology.ts:68

Tenant ID (overridden by X-Tenant-Id header when present).