agents/workflows/performance-optimization.md

model

model
claude-opus-4-1

Optimize application performance end-to-end using specialized performance and optimization agents:

[Extended thinking: This workflow coordinates multiple agents to identify and fix performance bottlenecks across the entire stack. From database queries to frontend rendering, each agent contributes their expertise to create a highly optimized application.]

Phase 1: Performance Analysis

1. Application Profiling

• Use Task tool with subagent_type="performance-engineer"
• Prompt: "Profile application performance for: $ARGUMENTS. Identify CPU, memory, and I/O bottlenecks. Include flame graphs, memory profiles, and resource utilization metrics."
• Output: Performance profile, bottleneck analysis, optimization priorities

2. Database Performance Analysis

• Use Task tool with subagent_type="database-optimizer"
• Prompt: "Analyze database performance for: $ARGUMENTS. Review query execution plans, identify slow queries, check indexing, and analyze connection pooling."
• Output: Query optimization report, index recommendations, schema improvements

Phase 2: Backend Optimization

3. Backend Code Optimization

• Use Task tool with subagent_type="performance-engineer"
• Prompt: "Optimize backend code for: $ARGUMENTS based on profiling results. Focus on algorithm efficiency, caching strategies, and async operations."
• Output: Optimized code, caching implementation, performance improvements

4. API Optimization

• Use Task tool with subagent_type="backend-architect"
• Prompt: "Optimize API design and implementation for: $ARGUMENTS. Consider pagination, response compression, field filtering, and batch operations."
• Output: Optimized API endpoints, GraphQL query optimization, response time improvements

Phase 3: Frontend Optimization

5. Frontend Performance

• Use Task tool with subagent_type="frontend-developer"
• Prompt: "Optimize frontend performance for: $ARGUMENTS. Focus on bundle size, lazy loading, code splitting, and rendering performance. Implement Core Web Vitals improvements."
• Output: Optimized bundles, lazy loading implementation, performance metrics

6. Mobile App Optimization

• Use Task tool with subagent_type="mobile-developer"
• Prompt: "Optimize mobile app performance for: $ARGUMENTS. Focus on startup time, memory usage, battery efficiency, and offline performance."
• Output: Optimized mobile code, reduced app size, improved battery life

Phase 4: Infrastructure Optimization

7. Cloud Infrastructure Optimization

• Use Task tool with subagent_type="cloud-architect"
• Prompt: "Optimize cloud infrastructure for: $ARGUMENTS. Review auto-scaling, instance types, CDN usage, and geographic distribution."
• Output: Infrastructure improvements, cost optimization, scaling strategy

8. Deployment Optimization

• Use Task tool with subagent_type="deployment-engineer"
• Prompt: "Optimize deployment and build processes for: $ARGUMENTS. Improve CI/CD performance, implement caching, and optimize container images."
• Output: Faster builds, optimized containers, improved deployment times

Phase 5: Monitoring and Validation

9. Performance Monitoring Setup

• Use Task tool with subagent_type="devops-troubleshooter"
• Prompt: "Set up comprehensive performance monitoring for: $ARGUMENTS. Include APM, real user monitoring, and custom performance metrics."
• Output: Monitoring dashboards, alert thresholds, SLO definitions

10. Performance Testing

• Use Task tool with subagent_type="test-automator"
• Prompt: "Create performance test suites for: $ARGUMENTS. Include load tests, stress tests, and performance regression tests."
• Output: Performance test suite, benchmark results, regression prevention

Coordination Notes

• Performance metrics guide optimization priorities
• Each optimization must be validated with measurements
• Consider trade-offs between different performance aspects
• Document all optimizations and their impact

Performance optimization target: $ARGUMENTS