Interface: StreamOptions

Defined in: packages/workflow-engine/src/ai/types.ts:449

Properties

abortSignal?

optional abortSignal?: AbortSignal

Defined in: packages/workflow-engine/src/ai/types.ts:455

Abort signal to cancel the AI SDK call (pass-through)

---

maxRetries?

optional maxRetries?: number

Defined in: packages/workflow-engine/src/ai/types.ts:453

Maximum number of retries for the AI SDK call (pass-through)

---

maxTokens?

optional maxTokens?: number

Defined in: packages/workflow-engine/src/ai/types.ts:451

---

onChunk?

optional onChunk?: (chunk) => void

Defined in: packages/workflow-engine/src/ai/types.ts:458

Parameters

chunk

string

Returns

void

---

onStepEnd?

optional onStepEnd?: GenerateTextOnStepEndCallback<NoInfer<ToolSet>, NoInfer<Context>>

Defined in: packages/workflow-engine/src/ai/types.ts:464

Callback fired when each step completes (for collecting tool results)

---

providerOptions?

optional providerOptions?: Record<string, Record<string, unknown>>

Defined in: packages/workflow-engine/src/ai/types.ts:471

Provider-specific options passed directly to the AI SDK call. Use this to control reasoning per call, e.g. { openrouter: { reasoning: { enabled: false } } } or { anthropic: { thinking: { type: "disabled" } } }.

---

stopWhen?

optional stopWhen?: Arrayable<StopCondition<NoInfer<ToolSet>, Context>>

Defined in: packages/workflow-engine/src/ai/types.ts:462

Condition to stop tool execution (e.g., stepCountIs(3))

---

temperature?

optional temperature?: number

Defined in: packages/workflow-engine/src/ai/types.ts:450

---

timeoutMs?

optional timeoutMs?: number

Defined in: packages/workflow-engine/src/ai/types.ts:457

Override the helper-level per-call timeout.

---

tools?

optional tools?: ToolSet

Defined in: packages/workflow-engine/src/ai/types.ts:460

Tool definitions for the model to use