qa

Automates headless-browser QA testing, visual crawls, and performance benchmarking against running web applications.

Updated May 3, 2026
One-click install
npx skills add https://github.com/Claudfather/clauDNA --skill qa-claudfather
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: qa
Source: https://github.com/Claudfather/clauDNA/tree/main/skills/qa
Command: npx skills add https://github.com/Claudfather/clauDNA --skill qa-claudfather

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, and includes references (resource) components.

What problem does it solve? Manually clicking through every page of a web app to find bugs, broken links, console errors, and responsive layout issues is slow and inconsistent. This Skill drives a headless browser to systematically test a running app, capture screenshot evidence, and file findings as GitHub issues. ## Core Features & Use Cases - Seven QA modes: test-and-fix, report-only, canary deploy monitoring, performance benchmarking, ad-hoc browsing, cookie import for authenticated testing, and an autonomous full-site visual crawl. - Visual crawl with evidence: discovers routes, screenshots each at desktop/tablet/mobile viewports, collects console errors, tests interactions, and compares observed styles against design tokens using ΔE2000 color distance. - Structured findings: classifies issues by category and priority (Critical through Low), groups related findings, and files them as GitHub issues via the publish workflow. - Use Case: After deploying a new frontend build, run a canary check to diff screenshots and Core Web Vitals against the pre-deploy baseline and get alerted on regressions over 20%. ## Quick Start Ask the agent to run a QA visual crawl of your local dev server and file any bugs it finds as GitHub issues.

Frequently Asked Questions about qa

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run an automated visual crawl of my website?

Invoke the qa skill in crawl mode with a target, such as --url https://app.example.com or --local to spin up a dev server. It discovers routes, screenshots each at three viewports, checks console errors and interactions, then files findings as GitHub issues.

How do I test authenticated pages with a headless browser?

Use the cookies mode first to import cookies from your real Chromium browser through a domain picker. Subsequent test, crawl, or browse commands then run in an authenticated session against protected pages.

What is the difference between test mode and report mode?

Test mode finds bugs and fixes them iteratively, writing a failing test, applying a fix, and committing atomically per bug. Report mode stops after discovery and delivers a structured bug list with screenshots and repro steps without modifying code.

Can the crawl run fully autonomously without prompts?

Yes, crawl mode supports --auto, which skips plan mode and confirmations, implies GitHub issue output, and requires --url or --local. It emits a structured JSON result summarizing routes crawled, screenshots, and issues created.

When should I not use this QA skill?

Use a design audit skill for subjective design critique, a frontend performance audit for code-level rendering issues, and a security audit for vulnerabilities. This skill focuses on runtime browser behavior, not static code analysis.