Interface: GuardrailOutputOptions

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:102

Options for output guardrail wrapping.

Properties

holdUntilFinal?

optional holdUntilFinal: boolean

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:120

Hold every TEXT_DELTA until the final guards have judged the whole reply. Allowed or flagged, the held deltas go out before the final chunk; blocked or sanitized, they are dropped, so the caller never sees text the guards refused. An actionable tool call closes the window: the text so far is judged as a final reply before it.

---

inputEvaluations?

optional inputEvaluations: GuardrailEvaluationResult[] | null

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:110

Input evaluations to attach to first output chunk

---

onVerdict()?

optional onVerdict: (verdict) => void | Promise<void>

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:123

Called once with the final verdict on the reply: blocked, sanitized, flagged or allowed. A throw is logged and ignored.

Parameters

verdict

GuardrailOutputVerdict

Returns

void | Promise<void>

---

personaId?

optional personaId: string

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:107

Persona ID for error chunks

---

ragSources?

optional ragSources: RagRetrievedChunk[]

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:113

RAG sources to thread through to output guardrails for grounding verification

---

streamId

streamId: string

Defined in: packages/agentos/src/safety/guardrails/guardrailDispatcher.ts:104

Stream identifier for error chunks