Observability

Four hooks report request lifecycle events. They are configured per provider slot and never leak between providers.

Hook contracts

Hook	Fires	Payload
`onRequestStart`	Immediately before each attempt is sent	`{ url, method, attempt }`
`onResponse`	After each attempt completes, success or failure	`{ url, method, attempt, durationMs }`
`onError`	When an attempt fails, before each retry wait, and once more when retries are exhausted	`(error: AniLinkError, context)` with `context = { url, method, attempt, code, status?, nextDelayMs? }`
`onRetry`	When a failed attempt is going to be retried	Same shape as `onError` with `nextDelayMs` set

attempt is 1-based. durationMs is the elapsed wall-clock time of the attempt, making onResponse the natural point for latency metrics.

Firing order

onError fires for every failed attempt. onRetry fires only when another attempt will follow.

Usage

typescript

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

const aniLink = new AniLink("token", {
    onRequestStart: ({ url, attempt }) => {
        console.log("start", attempt, url);
    },
    onResponse: ({ url, durationMs }) => {
        console.log(url, `${durationMs}ms`);
    },
    onError: (error, { attempt, code }) => {
        console.error("attempt", attempt, "failed", code, error.message);
    },
    onRetry: (_error, { attempt, nextDelayMs }) => {
        console.log("retrying after", nextDelayMs, "ms; next attempt", attempt);
    },
});

Hook isolation

Hooks belong to the provider slot where they are declared. A hook registered for AniList never observes MAL traffic and vice versa:

typescript

const aniLink = new AniLink({
    anilist: { authToken: "t", onResponse: ({ durationMs }) => metrics.record(durationMs) },
    mal: { accessToken: "m" }, // no hooks — MAL traffic is unobserved
});

Next steps

Retries & resilience — the retry loop these hooks observe.
Per-request options — scoping options to a single call.

Observability ​

Hook contracts ​

Firing order ​

Usage ​