sharaf/landing-page-audit
Use when the user wants a landing page audit, CRO review, marketing conversion review, copy audit, funnel alignment check, brand positioning critique, technical performance marketing audit, or prioritized issue report for a landing page, pricing page, signup page, lead magnet, or product page. Runs evidence-first context gathering, parallel specialist analysis, synthesis, and severity-ranked recommendations.
100
100%
Does it follow best practices?
Impact
100%
1.81xAverage score across 3 eval scenarios
Passed
No known issues
SKILL.mdskills/landing-page-audit/
- name:
- landing-page-audit
- description:
- Use when the user wants a landing page audit, CRO review, marketing conversion review, copy audit, funnel alignment check, brand positioning critique, technical performance marketing audit, or prioritized issue report for a landing page, pricing page, signup page, lead magnet, or product page. Runs evidence-first context gathering, parallel specialist analysis, synthesis, and severity-ranked recommendations.
- metadata:
- {"version":"0.2.23","source_domain":"marketing","source_sub_domains":"landing-pages, cro, ad-creative-copywriting, brand-strategy, competitive-intelligence, full-funnel-growth","research_date":"2026-03-22"}
Landing Page Audit
Reference files under references/ hold phase details; open only the files needed for the current phase or specialist lens.
Purpose
Audit a landing page from marketing, conversion, copy, brand, technical, and funnel perspectives. Produce a narrative report with prioritized issues, evidence, conversion impact, and concrete fixes.
Input
The user should provide a local page URL, for example
http://localhost:3000. If no URL is provided, inspect the repo for a dev or
start script and infer the default port. Ask only if the URL cannot be
determined.
First Actions
Build a factual context brief before any critique. Use commands like these when local repo access is available:
rg --files | rg '(^|/)(README|package.json|vite.config|next.config|astro.config|src|app|pages|components|content|copy|public)'
rg -n "utm|gtag|GTM|fbq|analytics|pixel|conversion|thank-you|success|pricing|testimonial|logo|signup|demo|trial" .Collect product, audience, page type, page URL, visible copy, structure, CTAs, forms, social proof, tracking, images, metadata, and unavailable evidence. See context-gathering.md for the full brief contract and fallback rules.
Workflow
| # | Phase | Use |
|---|---|---|
| 1 | Gather context from repo, local render, and source | context-gathering.md |
| 2 | Run five specialist lenses | specialist-agents.md |
| 3 | Deduplicate and synthesize findings | synthesis-report.md |
| 4 | Apply guardrails, page-type logic, and success checks | guardrails-decision-logic.md |
Use parallel specialist agents when the host environment allows it. If not, run the five lenses as separate passes and keep their findings distinct before synthesis. No intermediate files are required; pass the context brief inline.
Quick Triage
| Signal | Classification |
|---|---|
| Form plus offer/value exchange, minimal navigation | Lead-generation landing page |
| Pricing tiers or plan comparison | Pricing page, bottom funnel |
| Account creation form with minimal content | Signup page, bottom funnel |
| Gated ebook, webinar, template, quiz, or report | Lead magnet, top funnel |
| Feature showcase with demo or trial CTA | Product page, mid funnel |
If traffic source is unknown, default to a cold-traffic lens and state that assumption. If the product name and target audience cannot be determined from repo, page, source, or metadata, ask: "What does this product do and who is it for?"
Severity Scale
| Severity | Definition | Examples |
|---|---|---|
| Critical | Known to cause more than 30% conversion loss | No CTA above fold, page load over 5s, ad-to-page mismatch, competing primary CTAs |
| High | 10-30% likely conversion impact | Missing social proof, generic headline, distracting navigation, weak trust signals |
| Medium | Measurable but under 10% expected impact | Form field friction, weak microcopy, inconsistent tone, incomplete metadata |
| Low | Best-practice gap or marginal lift | Missing sticky mobile CTA, no video, stock-like imagery signal |
Use benchmark data only from the bundled checklists. Do not invent statistics, traffic share, or device mix. Severity thresholds are internal; never quote them as conversion-loss percentages.
Finding Contract
Every finding block must use this exact template:
Finding: [what was observed]
Evidence: [specific quote, element, DOM/source signal, or file reference]
Why it matters: [impact explanation using bundled benchmark data when available]
Suggested fix: [concrete, actionable recommendation]
Severity: [critical/high/medium/low]Final Report Contract
Default to the report template in synthesis-report.md. The final answer must include:
- Executive summary with strengths and overall conversion readiness
- Detailed analysis for Copy & Messaging, Conversion & UX, Brand & Positioning, Technical Performance, and Funnel Alignment
- Prioritized issue list sorted Critical, High, Medium, Low
- Specific "What's Working Well" section to prevent over-correction
- Evidence gaps and assumptions, especially missing live-page access or unknown traffic source
- Form findings with more than four fields must state the bundled benchmark: optimal form length is about 4 fields.
The detailed analysis must include at least three structured finding blocks. A prioritized table is useful, but it does not replace those evidence-bearing blocks.