CtrlK
BlogDocsLog inGet started
Tessl Logo

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

1.81x
Quality

100%

Does it follow best practices?

Impact

100%

1.81x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

criteria.jsonevals/scenario-2/

{
  "context": "The agent audits a Next.js pricing page from source code only, with a Google Ads copy file also present in the repo. The audit should classify the page type and funnel stage correctly, check ad-to-page cohesion using the ads directory, flag form field count, flag navigation presence, and produce a structured report following the skill's exact output contract.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Pricing page classification",
      "description": "Report identifies the page as a pricing page (or equivalent: plan comparison page, bottom-funnel page) — not a generic landing page or signup page",
      "max_score": 8
    },
    {
      "name": "Bottom-funnel stage stated",
      "description": "Report explicitly states that a pricing page maps to the bottom of the funnel (or bottom-funnel, purchase-intent, or hot-traffic audience temperature)",
      "max_score": 8
    },
    {
      "name": "Ad copy cohesion checked",
      "description": "Report references the ads directory (inputs/ads/ or ads/google-search.md) and compares the ad headline or promise against the pricing page headline or CTA",
      "max_score": 10
    },
    {
      "name": "Ad headline mismatch flagged",
      "description": "Report flags that the ad headlines ('Stop Remote Worker Breaches Now', 'Zero-Trust Security', 'Built for IT Security Managers') do not match the pricing page headline ('Simple, Transparent Pricing')",
      "max_score": 8
    },
    {
      "name": "Navigation presence flagged",
      "description": "Report flags that the full navigation bar (with links to Home, Features, About, Blog, Careers, etc.) is present on the pricing page, creating exit paths",
      "max_score": 8
    },
    {
      "name": "Form field count flagged",
      "description": "Report flags that the demo request form has more than 4 fields (it has 8 fields: first name, last name, email, company, company size, job title, phone, use case) and references the optimal ~4-field benchmark",
      "max_score": 8
    },
    {
      "name": "Generic CTA flagged",
      "description": "Report flags 'Submit' as the form submit button label and identifies it as a weak CTA (or references that generic submit labels reduce conversions)",
      "max_score": 6
    },
    {
      "name": "Finding blocks with exact labels",
      "description": "At least three findings use the exact five-label block: 'Finding:', 'Evidence:', 'Why it matters:', 'Suggested fix:', 'Severity:' (case-insensitive)",
      "max_score": 8
    },
    {
      "name": "Severity correctly applied",
      "description": "At least one finding is rated Critical or High — either the navigation presence, the ad-to-page mismatch, or the excessive form fields",
      "max_score": 6
    },
    {
      "name": "Source-only limitation stated",
      "description": "Report states that live page access was unavailable and analysis is based on source code (or equivalent)",
      "max_score": 5
    },
    {
      "name": "Cold-traffic or known-traffic assumption",
      "description": "Report states a traffic-source assumption — either cold traffic (if UTM tracking is absent) or references the Google Ads campaign as a known warm-traffic signal",
      "max_score": 5
    },
    {
      "name": "No fabricated statistics",
      "description": "Report does NOT cite statistics that are not present in the skill's bundled benchmarks — i.e., does NOT invent conversion percentages or benchmark numbers beyond those in the skill (e.g. does not cite '73% of users abandon forms with more than 5 fields' or other made-up numbers)",
      "max_score": 5
    },
    {
      "name": "What's Working Well section",
      "description": "Report includes a 'What's Working Well' section (or 'Strengths') with at least one specific strength backed by evidence from the source files",
      "max_score": 5
    },
    {
      "name": "Severity-sorted issue list",
      "description": "Report contains a prioritized issue table or list where Critical and High severity items appear before Medium and Low items",
      "max_score": 5
    },
    {
      "name": "No visual hallucinations",
      "description": "Report does NOT make claims about visual quality (e.g. 'the design looks clean', 'colors are well-chosen') that are not backed by specific CSS values, class names, or DOM structure evidence from the source files",
      "max_score": 5
    }
  ]
}

evals

README.md

tile.json