home / skills / openclaw / skills / lighthouse-fix

lighthouse-fix skill

/skills/lxgicstudios/lighthouse-fix

This skill runs Lighthouse audits and delivers AI-generated, actionable fixes tailored to your site's results to boost performance.

npx playbooks add skill openclaw/skills --skill lighthouse-fix

Review the files below or copy the command above to add this skill to your agents.

Files (8)
SKILL.md
2.0 KB
---
name: lighthouse-fixer
description: Run Lighthouse audit and get AI fix suggestions. Use when improving performance.
---

# Lighthouse Fixer

Lighthouse tells you what's wrong but the fix suggestions are generic. This tool runs Lighthouse and gives you specific, actionable fixes for your issues.

**One command. Zero config. Just works.**

## Quick Start

```bash
npx ai-lighthouse https://mysite.com
```

## What It Does

- Runs full Lighthouse audit on any URL
- Analyzes the results with AI
- Provides specific fixes, not generic advice
- Prioritizes issues by impact

## Usage Examples

```bash
# Audit your site
npx ai-lighthouse https://mysite.com

# Specific page
npx ai-lighthouse https://example.com/page

# Focus on performance
npx ai-lighthouse https://mysite.com --category performance
```

## Best Practices

- **Fix performance first** - biggest impact on user experience
- **Test on slow connections** - not everyone has gigabit
- **Check mobile separately** - mobile scores are often worse
- **Iterate** - fix one thing at a time

## When to Use This

- Lighthouse scores are tanking
- SEO audit flagged performance issues
- Core Web Vitals are failing
- Want actionable advice, not documentation links

## Part of the LXGIC Dev Toolkit

This is one of 110+ free developer tools built by LXGIC Studios. No paywalls, no sign-ups, no API keys on free tiers. Just tools that work.

**Find more:**
- GitHub: https://github.com/LXGIC-Studios
- Twitter: https://x.com/lxgicstudios
- Substack: https://lxgicstudios.substack.com
- Website: https://lxgicstudios.com

## Requirements

No install needed. Just run with npx. Node.js 18+ recommended. Needs OPENAI_API_KEY environment variable.

```bash
npx ai-lighthouse --help
```

## How It Works

Runs Lighthouse against your URL, captures the full report, and sends it to GPT-4o-mini. The AI interprets the findings and provides specific code-level recommendations based on the actual issues found.

## License

MIT. Free forever. Use it however you want.

Overview

This skill runs a full Lighthouse audit for any public URL and returns AI-powered, code-level fixes you can act on immediately. It replaces generic recommendations with prioritized, concrete steps that target your real issues. One command, no config: designed to speed up performance, SEO, and Core Web Vitals remediation.

How this skill works

The tool launches Lighthouse against the supplied URL, captures the complete audit report, and sends the findings to an LLM for analysis. The AI reviews failing audits, pinpoints root causes, and generates clear, prioritized remediation steps and code suggestions. You get actionable tasks instead of links to docs, with emphasis on impact and quick wins.

When to use it

  • When Lighthouse scores drop and you need targeted remediation
  • After an SEO audit flags performance or load-time issues
  • To investigate failing Core Web Vitals and mobile performance problems
  • Before releases to ensure performance regressions are caught
  • When you want prioritized, developer-friendly fixes rather than generic advice

Best practices

  • Fix high-impact performance issues first (e.g., render-blocking resources, large images)
  • Run audits on mobile and slow network settings to mirror real user conditions
  • Iterate: apply one change, re-run the audit, verify improvements
  • Integrate into CI for regression checks on key pages
  • Provide authenticated or staging URLs when applicable to test realistic pages

Example use cases

  • Audit the homepage to reduce Largest Contentful Paint (LCP) and receive specific image or CSS fixes
  • Check a product page with heavy third‑party scripts and get recommendations for deferring or removing them
  • Debug a mobile-specific layout shift and obtain actionable steps to stabilize layout
  • Run before deployment to catch regressions introduced by new assets or libraries
  • Prioritize backlog issues by impact and receive concrete implementation suggestions

FAQ

Do I need to install anything?

No local install is required; run it with npx and Node.js 18+. An OPENAI_API_KEY environment variable is needed.

Can it audit private or authenticated pages?

Public URLs work out of the box. For authenticated pages, use staging URLs or a test path that replicates the authenticated view.