AniList client configuration

Complete options table

Option	Type	Default	Behavior	When to change it
`timeout`	`number`	`30000`	Milliseconds before the client aborts a request. `0` disables. Timeout errors carry `timeoutMs`	Slow endpoints or strict latency budgets
`signal`	`AbortSignal`	none	Cancels in-flight requests	User-driven cancellation
`retry`	`boolean \| Partial<RetryPolicy>`	built-in policy	Automatic retries. `false` opts out. Partial policy merges over defaults	Non-idempotent workflows, custom backoff
`paceWithRateLimit`	`boolean`	`true`	Waits for window reset when remaining quota drops below `rateLimitFloor`	High-volume schedulers
`rateLimitFloor`	`number`	`1`	Remaining-quota threshold that triggers pacing. It must be a finite, non-negative integer. A value of `0` disables floor-based pacing. A defined but invalid value throws	Start pacing earlier than the last request
`ignorePaceDeadline`	`boolean`	`false`	Bypass the shared pacing deadline for this call	Urgent single requests during a rate-limited window
`circuitBreaker`	`{ threshold, cooldownMs }`	off	Fail fast with `CIRCUIT_OPEN_ERROR` after `threshold` consecutive failures until `cooldownMs` elapses. Each consecutive failed probe doubles the next cooldown, capped at 8×	Protect against sustained outages
`responseCache`	`ResponseCache`	off	Opt-in TTL cache for reads: `GET` requests and GraphQL query documents	Read-heavy traversals with repeated identical reads
`onError`	`OnErrorHandler`	none	Fires per failed attempt and when retries run out	Logging, metrics
`onRetry`	`OnErrorHandler`	none	Fires when the client retries a failed attempt	Retry telemetry
`onRequestStart`	`OnRequestStartHandler`	none	Fires immediately before each attempt	Request counting
`onResponse`	`OnResponseHandler`	none	Fires after each attempt with `durationMs` and `cacheHit?`	Latency metrics, cache hit/miss tracking
`onPace`	`OnPaceHandler`	none	Fires after a rate-limit pacing wait completes, before the client sends the request. An aborted wait emits the elapsed portion with `aborted: true`	Pacing telemetry
`onHookError`	`OnHookErrorHandler`	none	Fires when a lifecycle hook throws. Wire this to your logger in production. Without it, hook failures fall back to `console.warn`, which is unqueryable in serverless/structured-logging setups	Route hook failures to a logger
`onCircuitOpen`	`OnCircuitOpenHandler`	none	Fires when the circuit breaker trips	Breaker trip alerts
`onCircuitClose`	`OnCircuitCloseHandler`	none	Fires when the circuit breaker closes after a probe	Breaker recovery tracking
`refreshToken`	`string`	none	Stored refresh token. With `clientId` and `clientSecret`, it enables automatic refresh on `401` and bootstraps a client without `authToken` on its first auth-required call.	Automatic token refresh
`clientId`	`string`	none	AniList application client ID. With `refreshToken` and `clientSecret`, it enables automatic refresh.	Automatic token refresh
`clientSecret`	`string`	none	AniList application secret. AniList requires it for the refresh grant, so automatic refresh stays off until you set it.	Automatic token refresh
`onTokenRefresh`	`AniListTokenRefreshCallback`	none	Fires after each successful automatic refresh and receives the effective token response. Persist it synchronously.	Persisting refreshed tokens
`exposeRawAxiosError`	`boolean`	`false`	Attaches the raw Axios error as `rawAxiosError`/`cause` and redacts sensitive headers	Local debugging only

Note

exposeRawAxiosError: true redacts the Authorization, Cookie, and Proxy-Authorization headers to [REDACTED] in the attached raw error, so opting in for diagnostics does not leak bearer tokens into your logs.

Usage

typescript

import { AniLink } from "anilink-api-wrapper";

const aniLink = new AniLink("anilist-token", {
    timeout: 10_000,
    retry: { maxRetries: 3, baseDelayMs: 250 },
    paceWithRateLimit: true,
    onResponse: ({ url, durationMs }) => console.log(url, durationMs),
});

Each mechanism gets its own page: Retries & resilience, Observability, Cancellation & timeouts, Response cache. With refreshToken, clientId, and clientSecret all set, a 401 triggers an automatic refresh and a single replay. See AniList authentication.

Next steps

Querying data covers what the configured client can call.
Per-request options explains how to override instance defaults for a single call.

AniList client configuration ​

Complete options table ​

Usage ​

Next steps ​

AniList client configuration

Complete options table

Usage

Next steps