React Data Fetching logo

React Data Fetching

Organization
PatternsDev
react-data-fetching

Teaches modern React data fetching patterns with TanStack Query, SWR, and Suspense. Use when implementing caching, deduplication, optimistic updates, or parallel loading in React applications.

Overview

PublisherPatternsDev
Repositoryskills
Skill namereact-data-fetching
Stars
250
Forks
27
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by PatternsDev on GitHub. Read the source before you install it.

Installation

Install the React Data Fetching AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/PatternsDev/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/react/react-data-fetching .claude/skills/react-data-fetching
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable React Data Fetching in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads React Data Fetching on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so React Data Fetching is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

React Data Fetching Patterns

Table of Contents

Production-ready patterns for fetching, caching, and synchronizing server data in React applications. These patterns are framework-agnostic — they work whether you're using Vite + React Router, Next.js, Remix, or a custom setup.

When to Use

Reference these patterns when:

  • Adding data fetching to components
  • Replacing useEffect + fetch with a proper data layer
  • Implementing caching, deduplication, or optimistic updates
  • Debugging waterfall loading patterns
  • Choosing between data fetching libraries

Instructions

  • Apply these patterns during code generation, review, and refactoring. When you see fetch-in-effect without caching or deduplication, suggest the appropriate pattern.

Details

Overview

The most common performance problem in React apps is request waterfalls — sequential fetches that could run in parallel. The second most common problem is redundant fetches — multiple components fetching the same data independently. The patterns below address both, starting with the highest-impact fixes.


1. Parallelize Independent Fetches with Promise.all

Impact: CRITICAL — Eliminates sequential waterfalls for 2-10x improvement.

When multiple fetches have no dependencies on each other, run them concurrently.

Avoid — sequential (3 round trips):

typescript
async function loadDashboard() {
  const user = await fetchUser()
  const posts = await fetchPosts()
  const notifications = await fetchNotifications()
  return { user, posts, notifications }
}

Prefer — parallel (1 round trip):

typescript
async function loadDashboard() {
  const [user, posts, notifications] = await Promise.all([
    fetchUser(),
    fetchPosts(),
    fetchNotifications(),
  ])
  return { user, posts, notifications }
}

When fetches have partial dependencies (B depends on A, but C doesn't), start independent work immediately:

typescript
async function loadPage() {
  const userPromise = fetchUser()
  const configPromise = fetchConfig()

  const user = await userPromise
  const [config, posts] = await Promise.all([
    configPromise,
    fetchPosts(user.id), // depends on user
  ])
  return { user, config, posts }
}

2. Defer Await Until the Value Is Needed

Impact: HIGH — Starts work earlier without blocking on results you don't need yet.

A common mistake is to await each promise immediately, even when subsequent code doesn't need the result right away. Start the promise early, then await it at the point where you actually read the value.

Avoid — blocks unnecessarily:

typescript
async function loadProfile(userId: string) {
  const user = await fetchUser(userId)       // waits here
  const prefs = await fetchPreferences()     // starts only after user resolves
  const avatar = buildAvatarUrl(user.avatar)
  return { user, prefs, avatar }
}

Prefer — start early, await late:

typescript
async function loadProfile(userId: string) {
  const userPromise = fetchUser(userId)      // starts immediately
  const prefsPromise = fetchPreferences()    // starts immediately

  const user = await userPromise             // await when needed
  const avatar = buildAvatarUrl(user.avatar)
  const prefs = await prefsPromise           // may already be resolved

  return { user, prefs, avatar }
}

This is complementary to Promise.all — use defer-await when you need intermediate results between fetches, and Promise.all when you can wait for everything at once.


3. Use TanStack Query for Client-Side Data

Impact: CRITICAL — Automatic caching, deduplication, revalidation, and error handling.

Raw useEffect + fetch lacks caching, deduplication, retry, and background refresh. Use a data fetching library.

Avoid — no caching, no dedup, no error handling:

tsx
function UserProfile({ userId }: { userId: string }) {
  const [user, setUser] = useState<User | null>(null)
  const [loading, setLoading] = useState(true)

  useEffect(() => {
    setLoading(true)
    fetch(`/api/users/${userId}`)
      .then(r => r.json())
      .then(setUser)
      .finally(() => setLoading(false))
  }, [userId])

  if (loading) return <Skeleton />
  return <div>{user?.name}</div>
}

Prefer — TanStack Query (recommended for Vite + React apps):

tsx
import { useQuery } from '@tanstack/react-query'

function UserProfile({ userId }: { userId: string }) {
  const { data: user, isLoading } = useQuery({
    queryKey: ['user', userId],
    queryFn: () => fetch(`/api/users/${userId}`).then(r => r.json()),
  })

  if (isLoading) return <Skeleton />
  return <div>{user?.name}</div>
}

TanStack Query is the strongest choice for Vite apps — it's framework-agnostic, has built-in useSuspenseQuery, devtools, infinite queries, optimistic mutations, and offline support. SWR is a lighter alternative that covers the basics (dedup, caching, revalidation) but has fewer features for complex mutation workflows.

Both give you: request deduplication, stale-while-revalidate caching, automatic retries, and background refresh.

Setup for Vite apps:

tsx
// main.tsx
import { QueryClient, QueryClientProvider } from '@tanstack/react-query'

const queryClient = new QueryClient({
  defaultOptions: {
    queries: {
      staleTime: 60_000, // 1 minute
      retry: 2,
    },
  },
})

createRoot(document.getElementById('root')!).render(
  <QueryClientProvider client={queryClient}>
    <App />
  </QueryClientProvider>
)

4. Use Suspense for Declarative Loading States

Impact: HIGH — Cleaner code, automatic loading coordination, streaming support.

Suspense lets you declare loading boundaries in the component tree instead of managing isLoading state in every component.

Avoid — manual loading orchestration:

tsx
function Dashboard() {
  const { data: user, isLoading: userLoading } = useQuery(userQuery)
  const { data: stats, isLoading: statsLoading } = useQuery(statsQuery)

  if (userLoading || statsLoading) return <FullPageSpinner />
  return (
    <div>
      <UserHeader user={user} />
      <StatsPanel stats={stats} />
    </div>
  )
}

Prefer — Suspense boundaries:

tsx
function Dashboard() {
  return (
    <Suspense fallback={<FullPageSpinner />}>
      <DashboardContent />
    </Suspense>
  )
}

function DashboardContent() {
  const { data: user } = useSuspenseQuery(userQuery)
  const { data: stats } = useSuspenseQuery(statsQuery)
  return (
    <div>
      <UserHeader user={user} />
      <StatsPanel stats={stats} />
    </div>
  )
}

For independent sections, use separate Suspense boundaries so they load independently:

tsx
function Dashboard() {
  return (
    <div>
      <Suspense fallback={<HeaderSkeleton />}>
        <UserHeader />
      </Suspense>
      <Suspense fallback={<StatsSkeleton />}>
        <StatsPanel />
      </Suspense>
    </div>
  )
}

TanStack Query provides useSuspenseQuery and SWR provides { suspense: true } option.


5. Prefetch Data Before Navigation

Impact: HIGH — Eliminates loading states on page transitions.

Start fetching data before the user commits to a navigation — on hover, focus, or route preload.

With TanStack Query:

tsx
import { useQueryClient } from '@tanstack/react-query'

function ProjectLink({ projectId }: { projectId: string }) {
  const queryClient = useQueryClient()

  const prefetch = () => {
    queryClient.prefetchQuery({
      queryKey: ['project', projectId],
      queryFn: () => fetchProject(projectId),
      staleTime: 30_000,
    })
  }

  return (
    <Link
      to={`/projects/${projectId}`}
      onMouseEnter={prefetch}
      onFocus={prefetch}
    >
      View Project
    </Link>
  )
}

With React Router loaders (Vite apps):

tsx
// routes.tsx
const routes = [
  {
    path: '/projects/:id',
    loader: ({ params }) => queryClient.ensureQueryData({
      queryKey: ['project', params.id],
      queryFn: () => fetchProject(params.id!),
    }),
    Component: ProjectPage,
  },
]

6. Use React.cache() for Server-Side Deduplication

Impact: MEDIUM — Deduplicates expensive operations within a single server render.

In server components (RSC), React.cache() ensures the same async call made by multiple components only executes once per request.

typescript
import { cache } from 'react'

export const getSession = cache(async () => {
  const session = await auth()
  if (!session?.user?.id) return null
  return session
})

export const getUser = cache(async (userId: string) => {
  return await db.user.findUnique({ where: { id: userId } })
})

Multiple components calling getSession() in the same render share one execution.

Important: Use primitive arguments (strings, numbers) for cache keys. Inline objects create new references and cause cache misses:

typescript
// Cache miss every time — new object reference
getUser({ id: '123' })
getUser({ id: '123' }) // miss

// Cache hit — same string value
getUser('123')
getUser('123') // hit

7. Implement Optimistic Updates for Instant Feedback

Impact: HIGH — UI responds immediately without waiting for the server.

For mutations where the outcome is predictable (toggling a like, updating a name), update the UI instantly and reconcile with the server response.

With TanStack Query:

tsx
import { useMutation, useQueryClient } from '@tanstack/react-query'

function LikeButton({ postId }: { postId: string }) {
  const queryClient = useQueryClient()

  const { mutate: toggleLike } = useMutation({
    mutationFn: () => fetch(`/api/posts/${postId}/like`, { method: 'POST' }),
    onMutate: async () => {
      await queryClient.cancelQueries({ queryKey: ['post', postId] })
      const previous = queryClient.getQueryData<Post>(['post', postId])
      queryClient.setQueryData<Post>(['post', postId], old => ({
        ...old!,
        liked: !old!.liked,
        likeCount: old!.liked ? old!.likeCount - 1 : old!.likeCount + 1,
      }))
      return { previous }
    },
    onError: (_err, _vars, context) => {
      queryClient.setQueryData(['post', postId], context?.previous)
    },
    onSettled: () => {
      queryClient.invalidateQueries({ queryKey: ['post', postId] })
    },
  })

  return <button onClick={() => toggleLike()}>Like</button>
}

8. Avoid Fetch Waterfalls in Component Trees

Impact: CRITICAL — Parent-then-child fetching is the #1 performance problem.

When a parent fetches data and a child fetches its own data based on the parent's result, you create a waterfall. Restructure to fetch in parallel.

Avoid — child can't start until parent finishes:

tsx
function UserPage({ userId }: { userId: string }) {
  const { data: user } = useQuery({
    queryKey: ['user', userId],
    queryFn: () => fetchUser(userId),
  })

  if (!user) return <Skeleton />
  return <UserPosts userId={user.id} /> // starts fetching only after user loads
}

function UserPosts({ userId }: { userId: string }) {
  const { data: posts } = useQuery({
    queryKey: ['posts', userId],
    queryFn: () => fetchPosts(userId),
  })
  // ...
}

Prefer — fetch both at the same level:

tsx
function UserPage({ userId }: { userId: string }) {
  const { data: user } = useQuery({
    queryKey: ['user', userId],
    queryFn: () => fetchUser(userId),
  })
  const { data: posts } = useQuery({
    queryKey: ['posts', userId],
    queryFn: () => fetchPosts(userId),
  })

  if (!user) return <Skeleton />
  return (
    <div>
      <UserHeader user={user} />
      <PostList posts={posts ?? []} />
    </div>
  )
}

Or use a route-level loader to fetch all data before the component renders.


9. Deduplicate Global Event Listeners

Impact: MEDIUM — Prevents N listeners for N component instances.

When multiple component instances need the same global event (resize, scroll, online), share a single listener.

typescript
// hooks/useOnlineStatus.ts
import { useSyncExternalStore } from 'react'

function subscribe(callback: () => void) {
  window.addEventListener('online', callback)
  window.addEventListener('offline', callback)
  return () => {
    window.removeEventListener('online', callback)
    window.removeEventListener('offline', callback)
  }
}

function getSnapshot() {
  return navigator.onLine
}

export function useOnlineStatus() {
  return useSyncExternalStore(subscribe, getSnapshot, () => true)
}

useSyncExternalStore automatically deduplicates subscriptions and ensures consistent state across concurrent renders.


10. Use Passive Event Listeners for Scroll and Touch

Impact: LOW-MEDIUM — Prevents scroll jank from blocking listeners.

Non-passive scroll/touch listeners block the browser's compositor thread. Mark them passive when you don't call preventDefault().

Avoid — blocks scrolling:

tsx
useEffect(() => {
  const handler = () => trackScroll(window.scrollY)
  window.addEventListener('scroll', handler)
  return () => window.removeEventListener('scroll', handler)
}, [])

Prefer — non-blocking:

tsx
useEffect(() => {
  const handler = () => trackScroll(window.scrollY)
  window.addEventListener('scroll', handler, { passive: true })
  return () => window.removeEventListener('scroll', handler)
}, [])

11. Schema-Version Your Client Storage

Impact: LOW-MEDIUM — Prevents crashes from stale localStorage data.

When reading from localStorage or sessionStorage, stale data from a previous app version can crash your app. Add a schema version and validate.

Avoid — crashes on schema change:

tsx
const [prefs, setPrefs] = useState(() => {
  return JSON.parse(localStorage.getItem('prefs') || '{}')
})

Prefer — versioned with fallback:

tsx
const PREFS_VERSION = 2

const [prefs, setPrefs] = useState<Prefs>(() => {
  try {
    const raw = localStorage.getItem('prefs')
    if (!raw) return DEFAULT_PREFS
    const parsed = JSON.parse(raw)
    if (parsed._v !== PREFS_VERSION) return DEFAULT_PREFS
    return parsed
  } catch {
    return DEFAULT_PREFS
  }
})

// On save, include version
useEffect(() => {
  localStorage.setItem('prefs', JSON.stringify({ ...prefs, _v: PREFS_VERSION }))
}, [prefs])

Source

Patterns from patterns.dev — framework-agnostic React data fetching guidance for the broader web engineering community.

Frequently asked questions

What does the React Data Fetching AI skill do?

Teaches modern React data fetching patterns with TanStack Query, SWR, and Suspense. Use when implementing caching, deduplication, optimistic updates, or parallel loading in React applications.

Why use React Data Fetching on TypingMind?

Because you install it once and use it with any model. React Data Fetching is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install React Data Fetching in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/PatternsDev/skills/tree/main/react/react-data-fetching. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use React Data Fetching?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with React Data Fetching?

As many as you like. As long as a model supports skills, you can use React Data Fetching with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the React Data Fetching AI skill free?

Yes. It is published on GitHub by PatternsDev under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇