performance-engineer_skill

Overview

This skill helps optimize application performance across build times, runtime speed, bundle size, and resource usage. It guides a systematic workflow: measure, profile, prioritize hotspots, apply targeted optimizations, and validate results. Use it to diagnose slow builds, heavy bundles, excessive memory use, or inefficient rendering.

How this skill works

The skill inspects performance metrics and guides profiling to identify real bottlenecks rather than assumptions. It recommends concrete tactics—caching, parallelization, code-splitting, and incremental changes—and integrates monitoring to validate improvements. It maps optimizations to likely causes (build, runtime, memory, network) and suggests targeted fixes.

When to use it

• Application or UI feels sluggish for users
• Build times or CI pipelines are excessively long
• Bundle size causes slow load times or high bandwidth costs
• Memory or CPU usage is high in production
• Database queries or backend endpoints are slow or inconsistent

Best practices

• Measure and profile before making changes; use real workloads and representative devices
• Prioritize fixes by impact and cost; attack the largest bottlenecks first
• Run incremental, validated changes with performance tests and benchmarks
• Use caching and CDN strategies where appropriate; prefer cache-first for stable data
• Split bundles and use lazy loading to reduce initial payload; keep hot paths small
• Monitor continuously in staging and production to detect regressions early

Example use cases

• Reduce initial load time by analyzing bundle composition and implementing code-splitting
• Shorten CI build times by parallelizing tasks and caching build artifacts
• Lower runtime CPU/memory by profiling hot functions and optimizing algorithms
• Improve perceived performance by deferring non-critical rendering and applying lazy hydration
• Add caching and TTL strategies for expensive API calls to cut latency and backend load