performance-engineering

Diagnose application performance by measuring CPU, memory, latency percentiles, and throughput.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/Samuelca6399/AbsolutelySkilled --skill performance-engineering-samuelca6399
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: performance-engineering
Source: https://github.com/Samuelca6399/AbsolutelySkilled/tree/main/skills/performance-engineering
Command: npx skills add https://github.com/Samuelca6399/AbsolutelySkilled --skill performance-engineering-samuelca6399

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you diagnose and improve slow, inefficient, or memory-leaking applications by turning performance work into a measurable workflow rather than guesswork.

Core Features & Use Cases

  • CPU Profiling & Flame Graphs: Identify hot functions, long tasks, and event-loop bottlenecks using tools like 0x, clinic.js, and built-in profilers.
  • Memory Leak Investigation: Capture and compare heap snapshots to find objects that keep accumulating across a workload window.
  • Benchmarking & Optimization Validation: Run fair microbenchmarks and confirm improvements with percentiles (P50/P95/P99), budgets, and realistic load testing.

Quick Start

Ask for a diagnosis and next commands by saying: "Use performance-engineering to investigate why our P99 latency regressed after the last deploy and outline the exact profiling and benchmarking steps."

Frequently Asked Questions about performance-engineering

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find a memory leak using heap snapshots?

Memory leak investigation involves capturing and comparing heap snapshots across a workload window to isolate accumulating objects. This measure-first approach validates the leak before applying fixes.

How do I diagnose high P99 latency after a deployment?

Diagnose P99 latency regression by profiling CPU hot functions, analyzing event-loop blocking, and running fair microbenchmarks. Use percentile budgets and realistic load testing to validate improvements.

What is the best way to identify CPU bottlenecks with flame graphs?

CPU profiling with flame graphs identifies hot functions and long tasks using toolchains like 0x and clinic.js. This visualizes call stacks to locate event-loop bottlenecks and optimize execution paths.

How do I set up fair microbenchmarks to validate performance improvements?

Set up fair microbenchmarks by measuring throughput and latency percentiles (P50, P95, P99) across production-like workloads. This validates optimizations with empirical data rather than assumptions.

When do I need event-loop blocking analysis for my application?

Event-loop blocking analysis is needed when profiling reveals long tasks or hot functions degrading throughput. It isolates synchronous operations stalling the loop, guiding targeted CPU optimizations.

Can I reduce frontend bundle size and optimize database queries with this profiling approach?

Frontend bundle reduction and database query optimization fit this measure-first profiling approach. It diagnoses bottlenecks by applying profiling toolchains and decision rules that focus on validated improvements.