agentsclimarketplace

Browser analyze

Skill Aetheria-Labs1/storefront-skills/skills/browser-analyze

Use Codex Browser to analyze a URL for design extraction, CRO evidence, or competitor research. Use when a task provides a reference or storefront URL.From its SKILL.md

Install
npx -y skills add Aetheria-Labs1/storefront-skills --skill browser-analyze

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 8 stars8 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

5.6 KB, as published. Nobody here has run it

Browser-Powered Page Analysis

Use Codex Browser to deeply analyze web pages before generating or optimizing storefronts.

When to Use

  • User says "remix this", "build like this", "analyze this page"
  • User provides a competitor URL or reference site
  • User wants CRO audit of their own page
  • User wants design tokens extracted from a live site

Workflow

Step 1 — Navigate and Capture

Use Codex Browser to:

  1. Open the target URL
  2. Take a full-page screenshot (desktop viewport)
  3. Resize to mobile (375px width) and screenshot again
  4. Note the page title, meta description, and any structured data

Step 2 — DOM Inspection

Use Codex Browser to run read-only JavaScript inspection:

// Extract structural data
({
  headings: [...document.querySelectorAll('h1,h2,h3')].map(h => ({ tag: h.tagName, text: h.textContent.trim() })),
  ctas: [...document.querySelectorAll('button, a[class*="btn"], a[class*="cta"], [data-action]')].map(el => ({ text: el.textContent.trim(), href: el.href || '', classes: el.className })),
  images: [...document.querySelectorAll('img')].slice(0, 20).map(img => ({ src: img.src, alt: img.alt, width: img.naturalWidth })),
  colors: getComputedStyle(document.body).backgroundColor,
  fonts: getComputedStyle(document.body).fontFamily,
  sections: [...document.querySelectorAll('section, [class*="section"], main > div')].length
})

From this data, identify:

  • Section count and heading hierarchy
  • CTA buttons (text, color, position relative to fold)
  • Color palette from computed styles
  • Font families in use
  • Image sources and alt text quality
  • Interactive elements (carousels, tabs, accordions, video players)

Step 3 — Conversion Analysis (12-Point Audit)

Score each 0-10 based on screenshots + DOM data:

#DimensionKey Question
1Above-the-FoldCTA visible without scroll? Hero compelling?
2Message-MatchHeadline aligned with likely traffic source?
3CTA QualityBenefit-driven copy? Adequate contrast? Single primary?
4Social ProofStars/testimonials/logos present? Well-placed?
5Trust SignalsShipping/returns/guarantees visible?
6Urgency/ScarcityReal or manufactured? Appropriate for vertical?
7Price PsychologyAnchoring? Payment splitting? Bundle breaks?
8Product PositioningBenefits-led? Lifestyle context? Differentiation?
9Mobile UX48px tap targets? 16px fonts? Sticky CTA?
10Section OrderingAIDA-compliant? FAQ before CTA?
11Page SpeedImage formats? LCP candidate? Render-blocking?
12Anti-PatternsAutoplay? Carousel hero? Nav on LP? Competing CTAs?

Step 4 — Design Token Extraction

Compile from visual inspection + computed styles:

{
  "primary_color": "#...",
  "secondary_color": "#...",
  "background": "#...",
  "surface": "#...",
  "text_color": "#...",
  "font_heading": "...",
  "font_body": "...",
  "border_radius": "...px",
  "spacing_unit": "...px"
}

Step 5 — Output PAGE_ANALYSIS

Output this structured block:

{
  "source_url": "https://...",
  "page_type": "pdp|landing|collection|homepage|editorial",
  "vertical": "beauty|supplements|fashion|food|luxury|home|tech",
  "sections_found": [
    { "type": "hero", "has_cta": true, "above_fold": true },
    { "type": "social-proof", "subtype": "logo-bar" },
    ...
  ],
  "design_tokens": {
    "primary_color": "#...",
    "font_heading": "...",
    "font_body": "...",
    "border_radius": "8px",
    "spacing": "comfortable|tight|generous"
  },
  "conversion_score": 72,
  "strengths": [
    "Strong hero with benefit-driven headline",
    "Trust badges visible above fold"
  ],
  "weaknesses": [
    { "dimension": "mobile-ux", "issue": "CTA below fold on mobile", "impact": "critical" },
    { "dimension": "social-proof", "issue": "No testimonials with real names", "impact": "high" }
  ],
  "recommended_remix": {
    "sections": ["hero", "trust-bar", "benefits", "testimonials", "product-showcase", "faq", "cta"],
    "tactics": ["sticky-cta", "real-testimonials", "benefit-before-feature"],
    "avoid": ["autoplay", "carousel-hero", "generic-stock"]
  }
}

Return PAGE_ANALYSIS_INPUT. Do not generate source HTML or page assets here.


Fallback (No @Browser Available)

If @Browser is not available or not enabled:

  1. Ask for screenshots or use any user-supplied visual reference.
  2. Note the limitation: no DOM access, mobile viewport test, or interaction detection.
  3. Hand the available evidence to analyze-page or remix.

Reference Data

Conversion Benchmarks

MetricAverageTop 20%Top 10%
All Shopify1.4%>3.2%>4.7%
Mobile1.2%—>3.9%
Landing pages (2026)3.5-5.2%——

Critical Multipliers

  • Sticky CTA + above-fold CTA: +12% CVR
  • Real testimonials with names: +22% CVR
  • Autoplay video: -7% CVR
  • Personalized CTAs: +202% vs default
  • 1s speed improvement: +2% CVR

Anti-Patterns to Detect

  • Autoplay video (loses 7% CVR)
  • Rotating carousels (banner blindness)
  • Navigation links on landing pages (exit opportunities)
  • Multiple competing CTAs per viewport
  • Generic stock photography
  • CTA below mobile fold without sticky alternative

Optional Follow-Up

This skill can end after returning PAGE_ANALYSIS_INPUT. That evidence can later inform analyze-page, remix, or optimize if the user requests one of those outcomes.

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.