Architecture Design Skill

Create robust system architectures, technical specifications, and make informed architectural decisions.

Capabilities

• •System architecture design
• •Technical specification writing
• •Technology stack selection
• •Database schema design
• •API design
• •Microservices vs monolith decisions
• •Scalability planning
• •Performance optimization strategy

Architecture Patterns

For SaaS Applications

Frontend (Next.js 14)
  ↓ API Routes
Backend Services
  ↓ Database
PostgreSQL (Supabase)
  + Auth
  + Realtime
  + Storage

For Multi-Agent Systems

User Request
  ↓
Supervisor Agent (LangGraph)
  ↓
Specialized Agents (Parallel)
  ↓
State Persistence (Checkpoints)
  ↓
Aggregated Response

For Scraping Projects

Input (URLs)
  ↓
Playwright (Browser Automation)
  ↓
Data Extraction
  ↓
Transformation
  ↓
Storage (JSON/Database)

Design Process

1. Requirements Analysis

• •Understand user needs
• •Define success criteria
• •Identify constraints (cost, time, tech)
• •List non-functional requirements

2. Technology Selection

"Using Deep Research skill:
 Research best options for: {requirement}
 Compare: {option A} vs {option B} vs {option C}
 Consider: performance, cost, DX, community support
 Recommend: Best option with justification"

3. Architecture Documentation

Create: docs/ARCHITECTURE.md

# System Architecture

## Overview
[High-level description]

## Tech Stack
- Frontend: {technology} - {reason}
- Backend: {technology} - {reason}
- Database: {technology} - {reason}

## Architecture Diagram
```mermaid
graph TD
    A[User] --> B[Frontend]
    B --> C[API]
    C --> D[Database]

Key Design Decisions

1. •Decision: Why we chose X over Y
2. •Decision: Why we structured Z this way

Scalability Strategy

• •Current: Supports X users
• •Future: Can scale to Y users via Z approach

Security Considerations

• •Auth: {approach}
• •Data: {encryption, RLS}
• •API: {rate limiting, validation}

### 4. Database Schema Design
```sql
-- Core entities
CREATE TABLE users (...);
CREATE TABLE main_entity (...);

-- Relationships
CREATE TABLE user_entity_mapping (...);

-- Always enable RLS
ALTER TABLE main_entity ENABLE ROW LEVEL SECURITY;

Architecture Review Checklist

Before finalizing architecture:

• • Meets all functional requirements
• • Scalable to expected user load
• • Cost-effective for current stage
• • Security best practices followed
• • Developer experience is good
• • Deployment strategy defined
• • Monitoring/observability planned
• • Backup/disaster recovery considered

Common Architecture Decisions

Monolith vs Microservices

Use Monolith (95% of projects):

• •Faster development
• •Simpler deployment
• •Lower overhead
• •Better for MVP/early stage

Use Microservices:

• •Team >50 engineers
• •Clear domain boundaries
• •Need independent scaling
• •Proven product-market fit

Database Choice

Supabase (PostgreSQL + extras):

• •Auth built-in
• •RLS for security
• •Realtime subscriptions
• •Good free tier

Plain PostgreSQL:

• •More control
• •Lower cost at scale
• •Custom infrastructure

MongoDB:

• •Document-based data
• •Flexible schema
• •Horizontal scaling

Remember: Perfect is the enemy of good. Start simple, scale when needed!