benchmark

Compare PR builds against baselines to detect performance regressions.

Updated Apr 3, 2026
One-click install
npx skills add https://github.com/lyh2/gstack-deepseek --skill benchmark-lyh2
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: benchmark
Source: https://github.com/lyh2/gstack-deepseek/tree/main/benchmark
Command: npx skills add https://github.com/lyh2/gstack-deepseek --skill benchmark-lyh2

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Establishes reliable baselines and lightweight processes to detect performance regressions across web pages, enabling teams to catch slowdowns early and maintain cross-PR performance health.

Core Features & Use Cases

  • Baseline capture & comparison: Create baselines and compare current results against historical data for PRs and branches.
  • Comprehensive metrics: Collect TTFB, FCP, LCP, DOM timings, and resource timing data; produce actionable insights and trends.
  • Trend & regression reporting: Identify regressions over time, surface hotspots, and generate recommendations for performance improvements.

Quick Start

Run /benchmark <url> to start a performance audit with baseline comparison

Frequently Asked Questions about benchmark

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

FAQPage Schema
How do I detect performance regressions in web page load timings across pull requests?

You can detect performance regressions by establishing historical baselines and comparing PR builds against them. This process applies to page load timings, Core Web Vitals, and resource sizes to catch slowdowns early across URLs and branches.

What is the best way to automate Core Web Vitals trend analysis for my web application?

Automating Core Web Vitals trend analysis involves collecting timing data like TTFB, FCP, and LCP, then comparing current results against saved historical baselines. This approach surfaces hotspots and generates recommendations for performance improvements over time.

How do I set up a performance baseline comparison for my branches?

To set up a performance baseline comparison, run a performance audit on a target URL to collect timing data. The system saves these baselines locally, allowing you to compare current branch results against historical data to identify regressions or warnings.

Can I track DOM timings and resource sizes across multiple URLs without external dependencies?

Yes, you can track DOM timings and resource sizes across multiple URLs without dependencies. The system requires a browse daemon to collect timing data and saves all baseline reports locally, ensuring comprehensive metrics collection without external modules.

Why do I need a browse daemon to collect performance timing data for regression reporting?

A browse daemon is required to actively collect timing data like TTFB, FCP, and LCP from targeted URLs. This background process enables the system to establish baselines, compare PR builds, and generate actionable regression and warning outputs.

What limitations should I expect when tracking performance trends over time using local baselines?

When tracking performance trends over time using local baselines, you are limited by the local storage of reports in your project directory. This approach focuses on PR and branch comparisons rather than continuous production monitoring.