Performance and composition patterns for React 19 + Vite + Cloudflare Workers projects. Use as a checklist when writing new components, a review guide when auditing existing code, or a refactoring playbook when something feels slow or tangled.
Works with
AI-first code editor with Composer
Before installing skills in Cursor, ensure your development environment meets these requirements:
node --versionreact-patternsExecute the skills CLI command in your project's root directory to begin installation:
Fetches react-patterns from jezweb/claude-skills and configures it for Cursor.
The CLI shows a list of agents. Use arrow keys and space to select Cursor:
Confirm successful installation by checking the skill directory location:
Restart Cursor to activate react-patterns. Access via /react-patterns in your agent's command palette.
We perform automated surface-level scans (Gen AI Scanner, Socket, Snyk) during installation. These checks detect common vulnerabilities but do not guarantee complete security. Always review skill source code and verify the publisher's reputation before production use.
Skills execute code in your environment. Always review source, verify the publisher, and test in isolation before production.
Submit your Claude Code skill and start earning
Automate repetitive workflows and reduce manual effort
Example
Generate reports, summarize documents, draft communications
Save 3-5 hours per week on routine tasks
Learn new skills, understand complex topics, get expert guidance
Example
Explain concepts, provide examples, suggest learning resources
Accelerate learning and skill development by 2x
Enhance output quality through reviews, suggestions, and refinements
Example
Review drafts, suggest improvements, catch errors
Improve work quality by 30-40% with less effort
0
total installs
0
this week
697
GitHub stars
0
upvotes
Run in your terminal
0
installs
0
this week
697
stars
Performance and composition patterns for React 19 + Vite + Cloudflare Workers projects. Use as a checklist when writing new components, a review guide when auditing existing code, or a refactoring playbook when something feels slow or tangled.
Rules are ranked by impact. Fix CRITICAL issues before touching MEDIUM ones.
Sequential async calls where they could be parallel. The #1 performance killer.
| Pattern | Problem | Fix |
|---|---|---|
| Await in sequence | const a = await getA(); const b = await getB(); |
const [a, b] = await Promise.all([getA(), getB()]); |
| Fetch in child | Parent renders, then child fetches, then grandchild fetches | Hoist fetches to the highest common ancestor, pass data down |
| Suspense cascade | Multiple Suspense boundaries that resolve sequentially | One Suspense boundary wrapping all async siblings |
| Await before branch | const data = await fetch(); if (condition) { use(data); } |
Move await inside the branch — don't fetch what you might not use |
| Import then render | const Component = await import('./Heavy'); return <Component /> |
Use React.lazy() + <Suspense> — renders fallback instantly |
How to find them: Search for await in components. Each await is a potential waterfall. If two awaits are independent, they should be parallel.
Every KB the user downloads is a KB they wait for.
| Pattern | Problem | Fix |
|---|---|---|
| Barrel imports | import { Button } from '@/components' pulls the entire barrel file |
import { Button } from '@/components/ui/button' — direct import |
| No code splitting | Heavy component loaded on every page | React.lazy(() => import('./HeavyComponent')) + <Suspense> |
| Third-party at load | Analytics/tracking loaded before the app renders | Load after hydration: useEffect(() => { import('./analytics') }, []) |
| Full library import | import _ from 'lodash' (70KB) |
import debounce from 'lodash/debounce' (1KB) |
| Lucide tree-shaking | import * as Icons from 'lucide-react' (all icons) |
Explicit map: import { Home, Settings } from 'lucide-react' |
| Duplicate React | Library bundles its own React → "Cannot read properties of null" | resolve.dedupe: ['react', 'react-dom'] in vite.config.ts |
How to find them: npx vite-bundle-visualizer — shows what's in your bundle.
How you structure components matters more than how you optimise them.
| Pattern | Problem | Fix |
|---|---|---|
| Boolean prop explosion | <Card isCompact isClickable showBorder hasIcon isLoading> |
Explicit variants: <CompactCard>, <ClickableCard> |
| Compound components | Complex component with 15 props | Split into <Dialog>, <Dialog.Trigger>, <Dialog.Content> with shared context |
| renderX props | <Layout renderSidebar={...} renderHeader={...} renderFooter={...}> |
Use children + named slots: <Layout><Sidebar /><Header /></Layout> |
| Lift state | Sibling components can't share state | Move state to parent or context provider |
| Provider implementation | Consumer code knows about state management internals | Provider exposes interface { state, actions, meta } — implementation hidden |
| Inline components | function Parent() { function Child() { ... } return <Child /> } |
Define Child outside Parent — inline components remount on every render |
The test: If a component has more than 5 boolean props, it needs composition, not more props.
Not all re-renders are bad. Only fix re-renders that cause visible jank or wasted computation.
| Pattern | Problem | Fix |
|---|---|---|
| Default object/array props | function Foo({ items = [] }) → new array ref every render |
Hoist: const DEFAULT = []; function Foo({ items = DEFAULT }) |
| Derived state in effect | useEffect(() => setFiltered(items.filter(...)), [items]) |
Derive during render: const filtered = useMemo(() => items.filter(...), [items]) |
| Object dependency | useEffect(() => {...}, [config]) fires every render if config is {} |
Use primitive deps: useEffect(() => {...}, [config.id, config.type]) |
| Subscribe to unused state | Component reads { user, theme, settings } but only uses user |
Split context or use selector: useSyncExternalStore |
| State for transient values | const [mouseX, setMouseX] = useState(0) on mousemove |
Use useRef for values that change frequently but don't need re-render |
| Inline callback props | <Button onClick={() => doThing(id)} /> — new function every render |
useCallback or functional setState: <Button onClick={handleClick} /> |
How to find them: React DevTools Profiler → "Why did this render?" or <React.StrictMode> double-renders in dev.
Patterns that changed or are new in React 19.
| Pattern | Old (React 18) | New (React 19) |
|---|---|---|
| Form state | useFormState |
useActionState — renamed |
| Ref forwarding | forwardRef((props, ref) => ...) |
function Component({ ref, ...props }) — ref is a regular prop |
| Context | useContext(MyContext) |
use(MyContext) — works in conditionals and loops |
| Pending UI | Manual loading state | useTransition + startTransition for non-urgent updates |
| Route-level lazy | Works with createBrowserRouter only |
Still true — <Route lazy={...}> is silently ignored with <BrowserRouter> |
| Optimistic updates | Manual state management | useOptimistic hook |
| Metadata | Helmet or manual <head> management |
<title>, <meta>, <link> in component JSX — hoisted to <head> automatically |
| Pattern | Problem | Fix |
|---|---|---|
| Layout shift on load | Content jumps when async data arrives | Skeleton screens matching final layout dimensions |
| Animate SVG directly | Janky SVG animation | Wrap in <div>, animate the div instead |
| Large list rendering | 1000+ items in a table/list | @tanstack/react-virtual for virtualised rendering |
| content-visibility | Long scrollable content renders everything upfront | content-visibility: auto on off-screen sections |
| Conditional render with && | {count && <Items />} renders 0 when count is 0 |
Use ternary: {count > 0 ? <Items /> : null} |
| Pattern | Problem | Fix |
|---|---|---|
| No deduplication | Same data fetched by 3 components | TanStack Query or SWR — automatic dedup + caching |
| Fetch on mount | useEffect(() => { fetch(...) }, []) — waterfalls, no caching, no dedup |
TanStack Query: useQuery({ queryKey: ['users'], queryFn: fetchUsers }) |
| No optimistic update | User clicks save, waits 2 seconds, then sees change | useMutation with onMutate for instant visual feedback |
| Stale closure in interval | setInterval captures stale state |
useRef for the interval ID and current values |
| Polling without cleanup | setInterval in useEffect without clearInterval |
Return cleanup: useEffect(() => { const id = setInterval(...); return () => clearInterval(id); }) |
| Pattern | Problem | Fix |
|---|---|---|
import.meta.env in Node scripts |
Undefined — only works in Vite-processed files | Use loadEnv() from vite |
| React duplicate instance | Library bundles its own React | resolve.dedupe + optimizeDeps.include in vite.config.ts |
| Radix Select empty string | <SelectItem value=""> throws |
Use sentinel: <SelectItem value="__any__"> |
| React Hook Form null | {...field} passes null to Input |
Spread manually: value={field.value ?? ''} |
| Env vars at edge | process.env doesn't exist in Workers |
Use c.env (Hono context) or import.meta.env (Vite build-time) |
When reviewing code, go through categories 1-3 (CRITICAL + HIGH) for every PR. Categories 4-8 only when performance is a concern.
/react-patterns [file or component path]
Read the file, check against rules in priority order, report findings as:
file:line — [rule] description of issue
Prerequisites
Time Estimate
15-45 minutes depending on use case complexity
Steps
Common Pitfalls
✓ Do
✗ Don't
💡 Pro Tips
✓ Use when
Use when skill capabilities match your task, clear ROI on time saved, and you can validate outputs. Best for repetitive tasks, learning, and quality improvement.
✗ Avoid when
Avoid when task requires deep expertise you can't validate, involves sensitive decisions, or when learning process is more valuable than speed of completion.
jezweb/claude-skills
asyrafhussin/agent-skills
anthropics/claude-code
mblode/agent-skills
github/awesome-copilot
sickn33/antigravity-awesome-skills
Registry listing for react-patterns matched our evaluation — installs cleanly and behaves as described in the markdown.
Keeps context tight: react-patterns is the kind of skill you can hand to a new teammate without a long onboarding doc.
Registry listing for react-patterns matched our evaluation — installs cleanly and behaves as described in the markdown.
I recommend react-patterns for anyone iterating fast on agent tooling; clear intent and a small, reviewable surface area.
react-patterns reduced setup friction for our internal harness; good balance of opinion and flexibility.
I recommend react-patterns for anyone iterating fast on agent tooling; clear intent and a small, reviewable surface area.
Useful defaults in react-patterns — fewer surprises than typical one-off scripts, and it plays nicely with `npx skills` flows.
Keeps context tight: react-patterns is the kind of skill you can hand to a new teammate without a long onboarding doc.
react-patterns has been reliable in day-to-day use. Documentation quality is above average for community skills.
Useful defaults in react-patterns — fewer surprises than typical one-off scripts, and it plays nicely with `npx skills` flows.
showing 1-10 of 68