Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Frontend Visual QA

A rigorous visual QA protocol that audits already-rendered web, deck, dashboard, and Electron UIs with real-browser evidence and produces an evidence-backed defect report.

Dev & CodingAdvanced★ 1,323⑂ 212AI score 9/10Last updated: Aug 8, 2026

What it does

  • Audits the interface a user can actually see. Build success, DOM presence, and uninspected screenshot files count only as supporting signals, never as visual proof.
  • Forces a falsifiable audit contract first: target kind and canonical identity (URL, file hash, native artifact fingerprint), actor and job, conditional state (logged-out, empty, role-less, seeded), viewport matrix, design-system reference, and explicit authorization (audit-only vs fix-and-verify).
  • Ranks evidence A–D (real browser/native window, DevTools/E2E, headless sweep, static reasoning) and blocks promoting weak evidence into strong claims.
  • Ships scripts: scripts/visual_layout_audit.mjs (Playwright sweep for overflow, wrapping, media state, responsive widths) and scripts/silent_degradation_probe.mjs (font-never-shipped, renamed library classes, element crops).
  • Exercises real journeys — downloads, share/clipboard, file dialogs, print/PDF preview, Electron shell — then reports with severity (Blocker/Major/Moderate/Minor), crops, and an explicit "not verified" list.

Who it's for

  • Frontend developers doing pre-release visual review of web apps, landing pages, HTML decks, and admin dashboards.
  • Teams facing the "every gate is green but the page looks cheap" problem who need a detection method.
  • Electron/hybrid desktop developers verifying shell chrome, permissions, and file dialogs.
  • Leads who want consistent verified / partial / blocked QA reporting.

Examples

  1. "Audit the dashboard at localhost:5173 across desktop, tablet, and mobile" → prove the fixture state with --require, run the sweep, open every screenshot, then file findings.
  2. "Does deck.html clip text at 1920x1080 projection?" → run --file deck.html --page-type deck --viewport 1920x1080 and falsify apparent clipping against DOM geometry.
  3. "Verify the Export PDF button actually works" → trigger it in a real browser, confirm the print preview has non-blank pages, and mark anything unreachable as unverified.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/daymade/claude-code-skills/HEAD/frontend-visual-qa/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the frontend-visual-qa folder from the GitHub repo daymade/claude-code-skills into my ~/.claude/skills/frontend-visual-qa/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/daymade/claude-code-skills.git && mkdir -p ~/.claude/skills && cp -r claude-code-skills/frontend-visual-qa ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Clone the repository: git clone https://github.com/daymade/claude-code-skills.git
  3. Create the skills directory: mkdir -p ~/.claude/skills
  4. Copy this skill only: cp -r claude-code-skills/frontend-visual-qa ~/.claude/skills/
  5. (Recommended) Check that the project you plan to audit already has Playwright; if not, the Level C mechanical sweep is skipped and only Level A/B evidence is used.
  6. Restart Claude Code, then ask something like "run a visual QA pass on this page".
  7. Run the bundled scripts from inside the audited project so they resolve its Playwright: node ~/.claude/skills/frontend-visual-qa/scripts/visual_layout_audit.mjs --help
View source on GitHub ↗License: MIT