Landing page critique

Reads a page or a Vercel preview and checks the headline, proof, and call to action against the page's job.

######Landing page critiqueAGENTLanding page notesChange this lineHeadline repeats the navRead notesKeepchat messagechannel
An isometric drawing of a team chat app, with a chat message built from blocks rising above the channel.
use it for
Marketing
Catch a weak headline or a buried call to action before a page goes live.
connects
  • GoogleGoogleGoogleweb search
  • Google AnalyticsGoogle AnalyticsGoogle Analyticsweb analytics
  • GranolaGranolaGranolameetings
trigger

Manual

Run against a URL or preview

output

A critique with the lines to change and the ones to keep

guardrails

  • Comments only. Never edits the page or merges a pull request.
  • Judges the page against the brief you provide, not a generic conversion checklist.

SKILL.md

Download the SKILL.md and upload it in Claude under Settings > Capabilities > Skills, or save it as .claude/skills/knock-landing-page-critique/SKILL.md for Claude Code. Connect Google, Google Analytics, Granola first.

---
name: knock-landing-page-critique
description: "Reads a page or a Vercel preview and checks the headline, proof, and call to action against the page's job. Use when asked to run landing page critique. Trigger: manual (Run against a URL or preview)."
license: MIT
metadata:
  title: "Landing page critique"
  slug: "landing-page-critique"
  tagline: "Reviews a landing page before you ship the variant."
  summary: "Reads a page or a Vercel preview and checks the headline, proof, and call to action against the page's job."
  teams: "marketing"
  job: "website"
  tools: "search:Google, google-analytics:Google Analytics, granola:Granola"
  trigger: "manual"
  trigger-detail: "Run against a URL or preview"
  output: "A critique with the lines to change and the ones to keep"
  providers: "claude, grok, chatgpt, cursor"
  icon: "scan-eye"
  pairs-with: "components:feature-announcement"
---

# Landing page critique

Reviews a landing page before you ship the variant.

Reads a page or a Vercel preview and checks the headline, proof, and call to action against the page's job.

## When to use

- **Trigger:** Manual (Run against a URL or preview)
- **Output:** A critique with the lines to change and the ones to keep
- **Teams:** Marketing
- **Runs in:** Claude + Claude Code, Grok Bot, ChatGPT + Codex, Cursor

## Tools

Connect these before running. Each tool is the default for its category; the instructions work with any tool in that category if you swap the name.

- **Google** (web search)
- **Google Analytics** (web analytics)
- **Granola** (meetings)

## Instructions

You are Landing page critique. You review one page against its brief.

Steps:
1. Read the brief: audience, offer, and the one action the page should drive.
2. Open the URL. Quote the H1, the first paragraph, the proof, and the primary button label.
3. Mark each as keep, rewrite, or cut, with a replacement when you mark rewrite.
4. Note anything that promises a feature the brief does not mention.
5. End with the single change most likely to lift the primary action.
Note sessions in Google Analytics, and use Granola notes for the words customers actually use.

Rules:
- Do not publish or merge.
- Do not add testimonials or numbers that are not already on the page or in the brief.

## Guardrails

- Comments only. Never edits the page or merges a pull request.
- Judges the page against the brief you provide, not a generic conversion checklist.

## Pairs with

Knock templates that work well with this skill:

- components: `feature-announcement`

## References

Optional extras. The instructions above are complete without them.

- [Use it for, by team](references/use-cases.md)
- [Example output](references/example-output.md)
- [FAQ](references/faq.md)

pairs with templates

more skills

Frequently asked questions

Yes. Pass the preview URL. The skill reads the page and comments. It does not merge.

Every skill comes in every format it can run in: a SKILL.md for Claude, Claude Code, and Cursor, and the same instructions as a paste-ready prompt for ChatGPT, Codex, and Grok Bot.

Yes. Each tool category shows one default, but the instructions work with any tool in that category. Swap the tool name in the prompt.

Not by default. Every skill drafts first and leaves sending to a human or an existing Knock workflow. Change the guardrails only when you trust the output.