Does this skill automatically fix performance issues?

No, this skill is advisory-only. It identifies problems and recommends fixes but does not implement changes to your source code automatically.

What indicators does it use to detect project types?

It scans for file extensions like .jsx or .tsx for web projects and languages like Go, Rust, or Python for compute-heavy projects, as well as configuration files like package.json or vite.config.

What is the difference between web and compute reviews?

Web reviews focus on network latency, caching, and asset delivery, while compute reviews focus on algorithms, memory management, and CPU efficiency.

Can it handle full-stack projects?

Yes, it automatically detects both web and backend components and dispatches the appropriate reviewers for both domains simultaneously.

How do I apply the suggested fixes?

After reviewing the report, you can ask Claude to implement specific recommendations directly or use the /implement command with the performance review as context.

Performance Reviewer & Auditor

Name: Performance Reviewer & Auditor
Author: chrisallenlane

bychrisallenlane

•

Analíticas y Monitorización

Audits codebases for web and computational performance bottlenecks to identify optimization opportunities across the full stack.

This skill provides an advisory-only performance audit by automatically detecting whether a project involves web assets, backend logic, or both, and deploying specialized reviewer agents for each domain. It moves beyond intuition to identify structural issues in algorithms, memory usage, network latency, and asset delivery, providing a consolidated report with severity-ranked findings and cross-cutting architectural insights without modifying the source code.

Características Principales

01In-depth compute analysis for algorithmic efficiency, memory, and I/O

02Automatic project type detection for web, compute, or full-stack domains

03Parallel execution of specialized performance reviewer agents

04Consolidated reporting with cross-cutting synthesis of architectural bottlenecks

05Comprehensive web auditing including caching, asset delivery, and loading strategies

0612 GitHub stars

Casos de Uso

01Auditing a new codebase for performance bottlenecks before a major release

02Identifying the root cause of slow loading times or high CPU usage in complex applications

03Establishing a performance baseline and identifying missing benchmarking tooling

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add chrisallenlane/claude-swe-workflows review-perf

For use in Claude.ai and ChatGPT

Características Principales

01In-depth compute analysis for algorithmic efficiency, memory, and I/O

02Automatic project type detection for web, compute, or full-stack domains

03Parallel execution of specialized performance reviewer agents

04Consolidated reporting with cross-cutting synthesis of architectural bottlenecks

05Comprehensive web auditing including caching, asset delivery, and loading strategies

0612 GitHub stars

Casos de Uso

01Auditing a new codebase for performance bottlenecks before a major release

02Identifying the root cause of slow loading times or high CPU usage in complex applications

03Establishing a performance baseline and identifying missing benchmarking tooling

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add chrisallenlane/claude-swe-workflows review-perf

For use in Claude.ai and ChatGPT