agentleFS
buildwithreefbuildwithreef/claude-skills/skills/landing-page-critique/SKILL.md

landing-page-critique

Skill0 starsChanged 5 months agoMITView on GitHub

Delivers a blunt, structured teardown of a landing page — hero, offer clarity, trust signals, CTA, friction, and mobile — with scores and rewritten copy for everything that fails. Use when the user asks for a landing page review, critique, or teardown, or says things like "why isn't my page converting", "roast my landing page", or "review this page".

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No compliment sandwiches. Lead with the biggest problem.

    line 12
  2. Distinctive:

    Every criticism comes with the rewrite or the fix — actual copy, not "make it clearer."

    line 13
  3. Distinctive:

    Ask what the page is supposed to do (call?

    line 15
  4. Distinctive:

    2. Headline: Must name the outcome or the problem — not the company, not a welcome.

    line 28
  5. Distinctive:

    Don't critique brand aesthetics when conversion is the goal — ugly pages with clear offers outconvert pretty ones with vague offers.

    line 78
  6. Distinctive:

    Don't recommend A/B testing to a site with 200 visitors a month.

    line 79
  7. Distinctive:

    Don't pad the score to be nice.

    line 80

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Landing Page Critique

A useful teardown is specific, scored, and ships the fix. "The headline is weak" is a tweet; "the headline is weak, here's the stronger one" is a critique.

Ground rules

  • No compliment sandwiches. Lead with the biggest problem. If something's genuinely good, say so once and move on.
  • Every criticism comes with the rewrite or the fix — actual copy, not "make it clearer."
  • Score each area /10 and justify the number in one line.
  • Ask what the page is supposed to do (call? form? purchase?) and where traffic comes from before judging anything. A page can't fail a goal it doesn't have.

Intake

  • The page (URL, screenshot, or pasted copy — for URLs, review actual current content)
  • The one conversion goal
  • Traffic source (Google Ads, Facebook, organic, flyer QR code — this sets message-match expectations)
  • Who the customer is and how urgent their problem is

The teardown — nine areas, in order

1. The five-second test. Cover everything below the fold. From the hero alone: what is this, who's it for, what do I do next? If any answer takes thought, that's the first fix and probably the biggest one.

2. Headline. Must name the outcome or the problem — not the company, not a welcome. Rewrite formulas to reach for:

  • Outcome + timeframe: "A new furnace installed by Friday"
  • Problem you're rid of: "Never chase an invoice again"
  • Specific proof: "312 Des Moines kitchens remodeled" The vaguest word in the headline is the one to kill first ("solutions," "quality," "trusted").

3. Call to action. One primary CTA, repeated down the page — not three competing ones. Button text = verb + what happens: "Get my free quote" beats "Submit" beats "Learn more." Above the fold, visually loudest thing on the page. If the business runs on phone calls, the number is a CTA — clickable, huge on mobile.

4. Offer clarity. After 30 seconds a stranger should know: what exactly you get, what it costs (or how pricing works), what happens after you click. "Contact us to discuss your needs" is not an offer. Free quote, free inspection, $99 first visit, 10-minute callback — an offer has edges.

5. Trust. Reviews with full names and faces (or at least names and towns), review-platform logos with counts, licenses/certifications by number, guarantees with teeth ("on time or the service call is free"), photos of the actual team and actual work. Stock photos actively subtract trust — flag every one.

6. Friction. Count the form fields; every field past ~4 costs conversions — challenge each one ("do you need their address to give a quote?"). Check: does the form say what happens next? Is there a phone alternative? Any surprise steps (account creation, PDF download, "someone will reach out eventually")?

7. Visual hierarchy. Squint test: the eye should land on headline → proof → CTA. If everything is bold, nothing is. Walls of centered text, carousels, and auto-playing anything get flagged.

8. Mobile. Most local-business traffic is mobile. Check: hero readable without pinching, CTA/phone reachable with a thumb, form usable, load under ~3 seconds, no element wider than the screen.

9. Message match. The ad/post/QR that brought them must match the headline that greets them. Ad says "emergency AC repair," page says "Comfort solutions since 1987" → instant bounce. Score against the stated traffic source.

Output format

# Teardown: [page] — Overall [X]/10

**The one thing:** [the single change worth the most, in two sentences]

| Area | Score | Verdict in one line |
|---|---|---|
| 5-second test | x/10 | |
| Headline | x/10 | |
| CTA | x/10 | |
| Offer | x/10 | |
| Trust | x/10 | |
| Friction | x/10 | |
| Hierarchy | x/10 | |
| Mobile | x/10 | |
| Message match | x/10 | |

## Findings
[per area: **What's wrong** → **The fix** (with rewritten copy)]

## Do these three today
1. ...
2. ...
3. ...

What not to do

  • Don't critique brand aesthetics when conversion is the goal — ugly pages with clear offers outconvert pretty ones with vague offers.
  • Don't recommend A/B testing to a site with 200 visitors a month; they need bigger swings, not statistics.
  • Don't pad the score to be nice. A 4 that gets fixed beats a 7 that gets framed.

Built by Reef — AI systems for small business → https://buildwithreef.com

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.