agentleFS
claude-worldclaude-world/cf-browser/skills/visual-qa/SKILL.md

visual-qa

Skill9 starsChanged 8 months agoMITView on GitHub

Multi-viewport visual QA — screenshot pages at multiple device sizes and compare

20 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 20 edited versions we read keep these.

  • 3

    Distinctive

    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:

    Default viewports (user can override):

    line 26
  2. Distinctive:

    Read the saved PNG image

    line 49
  3. Distinctive:

    Check for visual issues:

    line 50

“Common” is measured: kept by at least 6 in 10 of the 20 edited versions of this file we could read. “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

Visual QA

Automated visual quality assurance using multi-viewport screenshots.

Prerequisites

CF Browser MCP server must be configured (see the CF Browser quick start).

Workflow

Step 1: Determine Target Pages

From user input, determine which pages to check:

  • Single URL: screenshot that page
  • Site URL: use browser_links to discover key pages (homepage, about, pricing, etc.), cap at 10
  • URL list: use provided list directly

Step 2: Define Viewports

Default viewports (user can override):

NameWidthHeightDevice
mobile375812iPhone 14
tablet7681024iPad
laptop128080013" laptop
desktop19201080Full HD monitor

Step 3: Capture Screenshots

For each page + viewport combination:

browser_screenshot(url, width=W, height=H)
→ Inspect the returned private temporary PNG
→ If persistent artifacts were requested, copy it to screenshots/{page-slug}/{viewport-name}.png before the MCP process exits

Process in batches to respect rate limits (4 viewports per page, pause between pages).

Step 4: Visual Review

For each screenshot:

  1. Read the saved PNG image
  2. Check for visual issues:
    • Layout broken or overlapping elements
    • Text overflow or truncation
    • Missing images or icons
    • Horizontal scrollbar on mobile
    • Navigation menu issues
    • Footer alignment

Step 5: Generate Report

Output a markdown report:

## Visual QA Report — {domain}
**Date**: {timestamp}
**Pages checked**: N
**Viewports**: mobile (375), tablet (768), laptop (1280), desktop (1920)

### Page: /
- mobile: OK
- tablet: WARNING — hero image overflows container
- laptop: OK
- desktop: OK

### Page: /pricing
- mobile: ISSUE — pricing table requires horizontal scroll
- tablet: OK
- laptop: OK
- desktop: OK

### Summary
- Total screenshots: N
- Issues found: M
- Warnings: W

Parameters

  • url (required): Page URL or site base URL
  • viewports (optional): Custom viewport list, default is mobile/tablet/laptop/desktop
  • pages (optional): Specific page paths to check
  • full_page (optional, default=false): Capture full scrollable page

Examples

"QA check our homepage on all devices"
→ 4 viewport screenshots of homepage + visual review

"Check the pricing page on mobile and desktop"
→ 2 viewport screenshots + review

"Visual QA the entire site after deploy"
→ Discover pages via browser_links → screenshot each at 4 viewports → report

"Full-page screenshot of the docs"
→ browser_screenshot(url, full_page=True) at each viewport

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.