techwavedev React Patterns
Principles for building production-ready React applications.
1. Component Design Principles
Component Types
| Type | Use | State |
|---|---|---|
| Server | Data fetching, static | None |
| Client | Interactivity | useState, effects |
| Presentational | UI display | Props only |
| Container | Logic/state | Heavy state |
Design Rules
- One responsibility per component
- Props down, events up
- Composition over inheritance
- Prefer small, focused components
2. Hook Patterns
When to Extract Hooks
| Pattern | Extract When |
|---|---|
| useLocalStorage | Same storage logic needed |
| useDebounce | Multiple debounced values |
| useFetch | Repeated fetch patterns |
| useForm | Complex form state |
Hook Rules
- Hooks at top level only
- Same order every render
- Custom hooks start with "use"
- Clean up effects on unmount
3. State Management Selection
| Complexity | Solution |
|---|---|
| Simple | useState, useReducer |
| Shared local | Context |
| Server state | React Query, SWR |
| Complex global | Zustand, Redux Toolkit |
State Placement
| Scope | Where |
|---|---|
| Single component | useState |
| Parent-child | Lift state up |
| Subtree | Context |
| App-wide | Global store |
4. React 19 Patterns
New Hooks
| Hook | Purpose |
|---|---|
| useActionState | Form submission state |
| useOptimistic | Optimistic UI updates |
| use | Read resources in render |
Compiler Benefits
- Automatic memoization
- Less manual useMemo/useCallback
- Focus on pure components
5. Composition Patterns
Compound Components
- Parent provides context
- Children consume context
- Flexible slot-based composition
- Example: Tabs, Accordion, Dropdown
Render Props vs Hooks
| Use Case | Prefer |
|---|---|
| Reusable logic | Custom hook |
| Render flexibility | Render props |
| Cross-cutting | Higher-order component |
6. Performance Principles
When to Optimize
| Signal | Action |
|---|---|
| Slow renders | Profile first |
| Large lists | Virtualize |
| Expensive calc | useMemo |
| Stable callbacks | useCallback |
Optimization Order
- Check if actually slow
- Profile with DevTools
- Identify bottleneck
- Apply targeted fix
7. Error Handling
Error Boundary Usage
| Scope | Placement |
|---|---|
| App-wide | Root level |
| Feature | Route/feature level |
| Component | Around risky component |
Error Recovery
- Show fallback UI
- Log error
- Offer retry option
- Preserve user data
8. TypeScript Patterns
Props Typing
| Pattern | Use |
|---|---|
| Interface | Component props |
| Type | Unions, complex |
| Generic | Reusable components |
Common Types
| Need | Type |
|---|---|
| Children | ReactNode |
| Event handler | MouseEventHandler |
| Ref | RefObject |
9. Testing Principles
| Level | Focus |
|---|---|
| Unit | Pure functions, hooks |
| Integration | Component behavior |
| E2E | User flows |
Test Priorities
- User-visible behavior
- Edge cases
- Error states
- Accessibility
10. Anti-Patterns
| ❌ Don't | ✅ Do |
|---|---|
| Prop drilling deep | Use context |
| Giant components | Split smaller |
| useEffect for everything | Server components |
| Premature optimization | Profile first |
| Index as key | Stable unique ID |
Remember: React is about composition. Build small, combine thoughtfully.
AGI Framework Integration
Qdrant Memory Integration
Before executing complex tasks with this skill:
python3 execution/memory_manager.py auto --query "<task summary>"
Decision Tree:
- Cache hit? Use cached response directly — no need to re-process.
- Memory match? Inject
context_chunksinto your reasoning. - No match? Proceed normally, then store results:
python3 execution/memory_manager.py store \
--content "Description of what was decided/solved" \
--type decision \
--tags react-patterns <relevant-tags>
Note: Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
Agent Team Collaboration
- Strategy: This skill communicates via the shared memory system.
- Orchestration: Invoked by
orchestratorvia intelligent routing. - Context Sharing: Always read previous agent outputs from memory before starting.
Local LLM Support
When available, use local Ollama models for embedding and lightweight inference:
- Embeddings:
nomic-embed-textvia Qdrant memory system - Lightweight analysis: Local models reduce API costs for repetitive patterns