Skip to content

Interface: ReasoningStageDto

@kortexya/reasoninglayer


@kortexya/reasoninglayer / ReasoningStageDto

Defined in: src/types/conversation.ts:521

One executed stage of the conversation pipeline (LLM-as-sensor trace).

optional confidence?: number

Defined in: src/types/conversation.ts:533

Optional confidence for this stage (0.0–1.0).


optional detail?: string

Defined in: src/types/conversation.ts:541

Free-text detail for display.


kind: string

Defined in: src/types/conversation.ts:527

“sensor” (LLM was consulted) or “symbolic” (deterministic).


llmUsed: boolean

Defined in: src/types/conversation.ts:531

Whether the LLM was actually invoked at this stage.


matchedEntities: MatchedEntityDto[]

Defined in: src/types/conversation.ts:535

Entities matched at this stage (sort + term id).


name: string

Defined in: src/types/conversation.ts:523

Stage id, e.g. “classify_intent”.


order: number

Defined in: src/types/conversation.ts:525

Execution order.


optional osfql?: string

Defined in: src/types/conversation.ts:537

OSFQL executed/produced at this stage, if any.


optional sensorPoint?: string

Defined in: src/types/conversation.ts:529

For sensor stages: the sensor point consulted.


optional strategy?: string

Defined in: src/types/conversation.ts:539

Symbolic strategy used, if any.