@okyrychenko-dev/react-effect-when

Declarative control over when React effects actually fire, especially when Strict Mode makes development noisy

react-effect-when helps you run effects only when dependencies reach the state you actually care about. Its main value is reducing development noise from React Strict Mode without disabling StrictMode, while also removing repeated useRef guards and if-based effect boilerplate.

What Problem It Solves

In real apps, many effects are not meant to run "on every mount-like moment". They should run only when something meaningful becomes true:

• a user and socket are both ready
• a modal is actually open
• analytics should fire once
• a subscription should start only after auth is available

In development, React Strict Mode can make those flows noisy by intentionally re-mounting and re-running effects. Teams often respond by:

• disabling Strict Mode
• adding ad hoc useRef(false) guards
• pushing conditional logic deep inside useEffect

react-effect-when gives you a cleaner option: keep Strict Mode on, but express effect timing declaratively.

Main Goals

• Reduce Strict Mode effect noise in development without turning Strict Mode off
• Replace repetitive useRef guards and early-return boilerplate with a declarative API
• Run effects only when dependencies are actually ready, truthy, or match a custom predicate
• Keep effect intent readable at the call site instead of hiding conditions inside the effect body
• Preserve predictable cleanup behavior and a familiar React mental model
• Provide strong TypeScript support for readiness and predicate-based narrowing

Why Use It

• Run an effect only when predicate(deps) becomes true
• Avoid repeating if (!user || !socket) return across components
• Reduce extra development noise around initialization, analytics, sockets, and one-time side effects
• Re-run only on meaningful matches with once: false
• Use useEffectWhenReady and useEffectWhenTruthy for common typed cases
• Keep public imports simple through the root package API

Installation

npm install @okyrychenko-dev/react-effect-when
# or
yarn add @okyrychenko-dev/react-effect-when
# or
pnpm add @okyrychenko-dev/react-effect-when

This package requires the following peer dependencies:

• React ^18.0.0 || ^19.0.0

Quick Start

import { useEffectWhenReady } from "@okyrychenko-dev/react-effect-when";

function Dashboard() {
  const [user, setUser] = useState<User | null>(null);
  const [socket, setSocket] = useState<Socket | null>(null);

  useEffectWhenReady(
    ([readyUser, readySocket]) => {
      readySocket.emit("identify", readyUser.id);
      return () => readySocket.emit("leave", readyUser.id);
    },
    [user, socket]
  );
}

Strict Mode In Development

This library is especially useful when you want development to feel closer to production for side effects that should not create extra noise.

Common examples:

• WebSocket or channel initialization
• analytics and tracking calls
• one-time setup effects
• notifications and toasts
• effects that should wait for fully ready data

The goal is not to fight React or replace useEffect. The goal is to make effect timing explicit and convenient in the cases where plain useEffect becomes noisy or repetitive.

Core Concepts

• useEffectWhen is the base hook. It receives the current dependency tuple and runs only when your predicate returns true.
• once: true means the effect runs once per mount lifecycle after the predicate first matches.
• once: false means the effect re-runs every time dependencies change and the predicate matches again.
• useEffectWhenReady is the fastest path when all dependencies must be non-null and non-undefined.
• useEffectWhenTruthy is the fastest path when all dependencies must be truthy.
• predicates.ready, predicates.truthy, and predicates.always are reusable building blocks for the base hook.

Core Use Cases

Reduce Strict Mode dev noise

Use useEffectWhen when a side effect should run only after a meaningful condition is satisfied instead of reacting to every mount-like development cycle.

Wait for async readiness

Use useEffectWhenReady when data, services, or refs load independently and the effect should wait until everything is available.

Remove repeated guard boilerplate

Use useEffectWhen when your current useEffect bodies mostly start with early returns and setup checks.

Re-run only on meaningful matches

Use once: false when you want the effect to run every time a threshold or condition is satisfied again.

Comparison

	Plain useEffect	Generic effect helper	@okyrychenko-dev/react-effect-when
Conditional effect execution	Manual guards inside the effect	Usually supported	Built-in
Wait for non-null async readiness	Manual guards	Varies	useEffectWhenReady
Wait for truthy values	Manual guards	Varies	useEffectWhenTruthy
Repeat only on meaningful matches	Manual branching	Varies	once: false
Access current deps tuple in the callback	Manual closure usage	Varies	Built-in
Observe skipped states	Manual logging	Rare	onSkip
Root-level simple public API	Native React only	Varies	Yes

Key Benefits

• Clear intent: the condition for running the effect is visible at the call site
• Less boilerplate: fewer manual refs, flags, and nested guards
• Better dev ergonomics: less Strict Mode noise without turning Strict Mode off
• Familiar semantics: still built on top of normal React effect behavior
• Typed readiness helpers: better safety when dependencies become available

When To Use It

• You want to reduce React Strict Mode double-invoke noise during development
• Your useEffect usually starts with guards like if (!user || !socket) return
• You would otherwise add useRef flags just to prevent effect running twice in development
• Your effect should wait until values are ready, truthy, or match a custom predicate
• You want cleanup behavior to stay explicit while the trigger condition stays readable

When Not To Use It

• A plain useEffect already expresses the behavior clearly
• The effect should always run for every dependency change with no gating
• The condition belongs in derived state or render logic rather than in an effect
• You are trying to bypass real remount semantics or "fix" React Strict Mode globally

Why Not Just Use useEffect + if + useRef

You can. For a one-off case, that is often fine.

The problem appears when the same pattern repeats across a codebase:

• if guards hide the real trigger condition inside the effect body
• useRef(false) flags add boilerplate and are easy to get wrong
• intent becomes inconsistent from component to component
• Strict Mode double invoke pain gets handled with ad hoc local workarounds

react-effect-when gives that pattern one explicit API instead of many custom versions.

Important Boundary

This library does not disable React Strict Mode, patch React behavior, or guarantee perfectly identical production behavior in every scenario.

What it does is make effect timing explicit and ergonomic in the cases where you want to reduce unnecessary development noise and avoid repeating local guard logic.

API Reference

Public imports

Use the root package import for all documented APIs:

import {
  predicates,
  useEffectWhen,
  useEffectWhenReady,
  useEffectWhenTruthy,
} from "@okyrychenko-dev/react-effect-when";

Hooks

Start with these first:

• useEffectWhen
• useEffectWhenReady
• useEffectWhenTruthy

useEffectWhen(effect, deps, predicate, options?)

Runs an effect only when predicate(deps) returns true.

Parameters:

• effect: (deps: T) => void | (() => void) - Same cleanup semantics as useEffect, plus access to the current dependency tuple
• deps: T extends DependencyList - Passed to React and to the predicate
• predicate: (deps: T) => boolean - Condition that controls when the effect runs
• options?: UseEffectWhenOptions<T>
  • once?: boolean - Run once after the first match or on every match
  • onSkip?: (deps: T) => void - Called when dependencies change and the predicate returns false; stops firing after the effect runs if once: true

Example:

import { useEffectWhen } from "@okyrychenko-dev/react-effect-when";

function Game({ score }: { score: number }) {
  useEffectWhen(
    ([currentScore]) => {
      showConfetti(currentScore);
    },
    [score],
    ([value]) => value > 100,
    { once: false }
  );
}

useEffectWhenReady(effect, deps, options?)

Runs the effect when all dependency values are non-null and non-undefined.

Parameters:

• effect: (deps: ReadyDeps<T>) => void | (() => void)
• deps: T extends DependencyList
• options?: UseEffectWhenOptions<T>

Example:

import { useEffectWhenReady } from "@okyrychenko-dev/react-effect-when";

function Profile({ user, token }: { user: User | null; token: string | null }) {
  useEffectWhenReady(
    ([readyUser, readyToken]) => {
      trackProfileView(readyUser.id, readyToken);
    },
    [user, token]
  );
}

useEffectWhenTruthy(effect, deps, options?)

Runs the effect when all dependency values are truthy.

Parameters:

• effect: (deps: TruthyDeps<T>) => void | (() => void)
• deps: T extends DependencyList
• options?: UseEffectWhenOptions<T>

Example:

import { useEffectWhenTruthy } from "@okyrychenko-dev/react-effect-when";

function SessionBanner({ token, isOnline }: { token: string | null; isOnline: boolean }) {
  useEffectWhenTruthy(
    ([readyToken, online]) => {
      connectBannerChannel(readyToken, online);
    },
    [token, isOnline],
    { once: false }
  );
}

Built-in predicates

• predicates.ready(deps) - true when all dependencies are non-null and non-undefined
• predicates.truthy(deps) - true when all dependencies are truthy
• predicates.always(deps) - always true, equivalent to a plain useEffect

Example:

import { predicates, useEffectWhen } from "@okyrychenko-dev/react-effect-when";

useEffectWhen(
  ([currentUser, currentToken]) => {
    initializeDashboard(currentUser, currentToken);
  },
  [user, token],
  predicates.ready,
  {
    onSkip: ([pendingUser, pendingToken]) => {
      console.debug("Waiting for deps:", {
        user: pendingUser,
        token: pendingToken,
      });
    },
  }
);

Real-World Examples

Prevent analytics from firing twice in development

import { useEffectWhen } from "@okyrychenko-dev/react-effect-when";

function ProductPage({ productId, isReady }: { productId: string; isReady: boolean }) {
  useEffectWhen(
    ([id]) => {
      analytics.track("product_view", { productId: id });
    },
    [productId, isReady],
    ([, ready]) => ready === true
  );
}

This is a common React Strict Mode double-invoke pain point in development when analytics should fire only after the page is actually ready.

Initialize a WebSocket only when auth is ready

import { useEffectWhenReady } from "@okyrychenko-dev/react-effect-when";

function RealtimeConnection({
  userId,
  authToken,
}: {
  userId: string | null;
  authToken: string | null;
}) {
  useEffectWhenReady(
    ([readyUserId, readyToken]) => {
      const socket = connectSocket({ userId: readyUserId, token: readyToken });

      return () => {
        socket.close();
      };
    },
    [userId, authToken]
  );
}

This keeps WebSocket setup declarative and avoids scattering if (!userId || !authToken) return checks through the effect body.

Show a toast only when a modal actually opens

import { useEffectWhen } from "@okyrychenko-dev/react-effect-when";

function Modal({ isOpen }: { isOpen: boolean }) {
  useEffectWhen(
    ([open]) => {
      toast.info("Modal opened");
    },
    [isOpen],
    ([open]) => open === true
  );
}

This is useful when development re-mounts would otherwise create extra toast or notification noise.

More Examples

Run once when a modal opens

import { useEffectWhen } from "@okyrychenko-dev/react-effect-when";

function Modal({ isOpen }: { isOpen: boolean }) {
  useEffectWhen(
    ([open]) => {
      fetchModalData(open);
    },
    [isOpen],
    ([open]) => open === true
  );
}

Sync only when a custom predicate matches

import { useEffectWhen } from "@okyrychenko-dev/react-effect-when";

useEffectWhen(
  ([itemList]) => {
    syncToServer(itemList);
  },
  [items, isOnline, hasPermission],
  ([itemList, online, permission]) => online === true && permission === true && itemList.length > 0
);

Observe skipped states with onSkip

import { predicates, useEffectWhen } from "@okyrychenko-dev/react-effect-when";

useEffectWhen(
  ([currentUser, currentToken]) => {
    initializeDashboard(currentUser, currentToken);
  },
  [user, token],
  predicates.ready,
  {
    onSkip: ([pendingUser, pendingToken]) => {
      console.debug("Still waiting:", {
        user: pendingUser,
        token: pendingToken,
      });
    },
  }
);

Internal Demo

The repository also contains a few internal demo snippets:

• examples/basic-ready.tsx
• examples/custom-predicate.tsx
• examples/truthy-repeat.tsx

Use them for local experimentation only. The README examples are the canonical public documentation for package usage.

Behavior Reference

Scenario	Behavior
Predicate false on mount	Effect does not run
Predicate becomes true	Effect runs
Deps change after effect ran (once: true)	Effect does not re-run
Deps toggle back to falsy, then truthy (once: true)	Effect does not re-run
Predicate true again (once: false)	Effect re-runs and the previous cleanup runs first
Component unmounts	Cleanup runs once
React 18 Strict Mode remount in dev	Same behavior as a fresh mount

Important Semantics

• predicate and onSkip are stored in refs, so the hook always uses their latest version without adding them to the dependency array.
• once is also kept fresh internally, so cleanup logic follows the latest option value across renders.
• Only deps control when React re-runs the effect. Changing predicate or onSkip alone does not trigger a re-run.
• useEffectWhen passes the current dependency tuple into effect.
• useEffectWhenReady and useEffectWhenTruthy are the preferred APIs when you want narrowed values for the common built-in conditions.
• In React Strict Mode, behavior is still scoped per mount lifecycle. A real remount is treated as a fresh hook instance.
• Public imports are exposed from the package root. Subpath imports are not required for the documented API.

Publish Checklist

Before publishing a new version, make sure this command passes:

npm run release:check

License

MIT © Oleksii Kyrychenko