Skill

performance-profiling

CPU/memory profiling, identifying bottlenecks, flame graphs, hotspot analysis.

Install

Run in your terminal

npx claudepluginhub sethdford/claude-skills --plugin engineer-debugging

Tool Access

This skill uses the workspace's default tool permissions.

Skill Content

Similar Skills

cache-components

Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.

cache-components

138.6k

claude-opus-4-5-migration

2 files

Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.

claude-opus-4-5-migration

83.2k

evaluation-methodology

1 file

Details PluginEval's skill quality evaluation: 3 layers (static, LLM judge), 10 dimensions, rubrics, formulas, anti-patterns, badges. Use to interpret scores, improve triggering, calibrate thresholds.

plugin-eval

32.9k

Stats

Parent Repo Stars3

Parent Repo Forks0

Last CommitMar 11, 2026

Actions

View Source View Plugin View on GitHub View README

Domain Context

CPU Profiling: Where does code spend time? Stack sampling at regular intervals

Memory Profiling: What allocates memory? Which objects aren't freed?

Flame Graph: Visual representation; width = time, stacked = call stack

Hot Spots: Functions consuming > 1% of time

Allocation Profiling: Which code allocates the most objects?

Instructions

Start Profiler: CPU profiler (perf, pprof, Java Flight Recorder)

Run Under Load: Profile should see realistic workload

Collect Data: Run for 10-30 seconds; enough to gather signal

Analyze Output: Flame graph or profiler GUI

Identify Hot Spots: Functions consuming most time

Drill Down: Zoom into hot functions; what calls them?

Measure Impact: Change code, re-profile, compare

Anti-Patterns

Profiling single-threaded code then deploying multi-threaded; lock contention changes profiles

Optimizing without profiling; "optimization" is just guessing

Micro-optimizing non-hotspot code; spend time on the hottest functions

Not considering cache effects; CPU cache profilers show real impact

Benchmarking without warm-up; first runs hit different code paths

Performance Profiling

Finding where code spends time and memory.

Context

You are profiling code to find bottlenecks. Use CPU and memory profilers.

Domain Context

CPU Profiling: Where does code spend time? Stack sampling at regular intervals
Memory Profiling: What allocates memory? Which objects aren't freed?
Flame Graph: Visual representation; width = time, stacked = call stack
Hot Spots: Functions consuming > 1% of time
Allocation Profiling: Which code allocates the most objects?

Instructions

Start Profiler: CPU profiler (perf, pprof, Java Flight Recorder)
Run Under Load: Profile should see realistic workload
Collect Data: Run for 10-30 seconds; enough to gather signal
Analyze Output: Flame graph or profiler GUI
Identify Hot Spots: Functions consuming most time
Drill Down: Zoom into hot functions; what calls them?
Measure Impact: Change code, re-profile, compare

Anti-Patterns

Profiling single-threaded code then deploying multi-threaded; lock contention changes profiles
Optimizing without profiling; "optimization" is just guessing
Micro-optimizing non-hotspot code; spend time on the hottest functions
Not considering cache effects; CPU cache profilers show real impact
Benchmarking without warm-up; first runs hit different code paths

performance-profiling

performance-profiling

Performance Profiling

Context

Domain Context

Instructions

Anti-Patterns

Further Reading

Performance Profiling

Context

Domain Context

Instructions

Anti-Patterns

Further Reading