RetryConfig

Defined in: core/retry.ts:23

Configuration options for retry behavior.

Example

// Custom retry with monitoring
const agent = LLMist.createAgent()
  .withRetry({
    retries: 5,
    minTimeout: 2000,
    onRetry: (error, attempt) => console.log(`Retry ${attempt}`),
  })
  .ask("Hello");

Properties

enabled?

optional enabled: boolean

Defined in: core/retry.ts:28

Whether retry is enabled.

Default

true

---

factor?

optional factor: number

Defined in: core/retry.ts:52

Exponential factor for backoff calculation.

Default

2

---

maxTimeout?

optional maxTimeout: number

Defined in: core/retry.ts:46

Maximum delay between retries in milliseconds.

Default

30000

---

minTimeout?

optional minTimeout: number

Defined in: core/retry.ts:40

Minimum delay before the first retry in milliseconds.

Default

1000

---

onRetriesExhausted()?

optional onRetriesExhausted: (error, attempts) => void

Defined in: core/retry.ts:70

Called when all retries are exhausted and the operation fails. The error will still be thrown after this callback.

Parameters

error

Error

attempts

number

Returns

void

---

onRetry()?

optional onRetry: (error, attempt) => void

Defined in: core/retry.ts:64

Called before each retry attempt. Use for logging or metrics.

Parameters

error

Error

attempt

number

Returns

void

---

randomize?

optional randomize: boolean

Defined in: core/retry.ts:58

Whether to add random jitter to prevent thundering herd.

Default

true

---

retries?

optional retries: number

Defined in: core/retry.ts:34

Maximum number of retry attempts.

Default

3

---

shouldRetry()?

optional shouldRetry: (error) => boolean

Defined in: core/retry.ts:78

Custom function to determine if an error should trigger a retry. If not provided, uses the default isRetryableError classification.

Parameters

error

Error

Returns

boolean

true to retry, false to fail immediately