Does it support the latest React features?

Yes, it includes specialized rules for React 19, the React Compiler for automatic memoization, and TanStack Virtual for high-performance list rendering.

What LLM performance features are included?

The skill covers vLLM deployment, quantization methods like AWQ and FP8, speculative decoding to increase throughput, and prompt caching to reduce token costs.

Can I use this for backend performance?

Absolutely. The skill includes rules for backend profiling using tools like py-spy and cProfile, along with Redis caching strategies and database optimization patterns.

How does this skill help with Core Web Vitals?

It provides specific implementation patterns for LCP, INP, and CLS, including resource preloading, layout shift prevention, and main-thread optimization using scheduler.yield.

Full-Stack Performance Optimization

Name: Full-Stack Performance Optimization
Author: yonatangross

byyonatangross

•

116

•

Analytics & Monitoring

Optimizes application speed across frontend Core Web Vitals, React rendering, backend profiling, and LLM inference efficiency.

This skill provides a comprehensive library of production-ready patterns for maximizing application performance and efficiency. It covers critical frontend metrics like Core Web Vitals (LCP, INP, CLS) using 2026 thresholds, React 19 render optimization with the new compiler, and advanced asset management through lazy loading and AVIF image optimization. Beyond the browser, it offers deep technical guidance on backend profiling, high-performance LLM inference using vLLM and quantization, and digital sustainability practices to reduce your application's carbon footprint.

Key Features

01116 GitHub stars

02Core Web Vitals optimization for LCP, INP, and CLS with future-proof 2026 thresholds

03Digital sustainability tools to enforce page weight budgets and reduce energy consumption

04Modern asset strategies for code-splitting, route-splitting, and AVIF/WebP image delivery

05High-performance LLM inference patterns featuring vLLM, quantization, and prompt caching

06React 19 render performance including automatic memoization and large-list virtualization

Use Cases

01Reducing AI infrastructure costs and latency through efficient LLM inference and caching

02Scaling React applications with large datasets using TanStack Virtual and the React Compiler

03Improving Google Search rankings by optimizing Core Web Vitals and Lighthouse scores

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yonatangross/orchestkit performance

For use in Claude.ai and ChatGPT

Download Skill