bench

Automate benchmarks and profiling to identify performance bottlenecks across commits.

Updated Jul 5, 2025
One-click install
npx skills add https://github.com/sqfzy/my_configs --skill bench-sqfzy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bench
Source: https://github.com/sqfzy/my_configs/tree/main/claude/skills/bench
Command: npx skills add https://github.com/sqfzy/my_configs --skill bench-sqfzy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates performance analysis workflows by running benchmarks, profiling, and comparing results to reveal bottlenecks and guide improvements.

Core Features & Use Cases

  • Execute baselines to quantify performance, then profile hotspots to locate slow paths.
  • Compare current results against baselines or specific commits to track progress.
  • Generate structured reports and artifacts that support reproducible performance improvements.

Quick Start

Run bench in profile mode against your target to generate a performance report.

Frequently Asked Questions about bench

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

FAQPage Schema
How do I identify performance bottlenecks in my codebase?

To identify performance bottlenecks, you can automate benchmarks and profiling tasks against your target codebase. This process profiles hotspots to locate slow paths and generates a structured report for reproducible improvements.

How do I benchmark code changes against a baseline commit?

Benchmarking against a baseline commit involves executing baselines to quantify performance, then comparing current results against that specific commit. This tracks progress and generates structured artifacts for reproducible performance analysis.

Does automated profiling support reproducible performance results?

Automated profiling supports reproducible performance results by recording the environment and enforcing deterministic benchmarks. This ensures structured reports accurately reflect code changes rather than environmental noise.

What is the best way to profile hotspots and locate slow code paths?

The best way to profile hotspots is running an automated profiling mode against your target codebase. This directly identifies performance bottlenecks and generates reports guiding structured optimization workflows.

Can I compare performance optimization workflows across different branches?

You can compare performance optimization workflows across different branches by comparing current benchmark results against baselines or specific commits. This tracks progress and validates that optimizations yield reproducible improvements.