GitHub - browser-use/qa-use

qa use

AI-Powered E2E Testing Platform ✨

A comprehensive, production-ready QA testing solution powered by BrowserUse AI agents. Automate your web application testing with intelligent agents that think, act, and validate like human testersβ€”but faster, more consistently, and around the clock.

Docker BrowserUse Next.js TypeScript

Test Run in Progress Test Run in Progress

πŸš€ Quick Start Guide

Prerequisites

  • 🐳 Docker & Docker Compose installed
  • πŸ”‘ BrowserUse API Key (get yours at cloud.browser-use.com)
  • πŸ“§ Resend API Key (optional, for email notifications)

⚑ 3-Step Setup

# 1️⃣ Clone the repository
git clone https://github.com/browser-use/qa-use.git
cd qa-use

# 2️⃣ Configure environment variables
cp .env.example .env
# Edit .env and add your API keys

# 3️⃣ Launch the platform
docker compose up

✨ Key Features & Highlights

  1. πŸ€– AI-Powered Testing Engine - Leverages BrowserUse to evaluate test cases in plain English.
  2. 🎯 Advanced Test Management - Organize tests into suites and run multiple tests in parallel.
  3. ⏰ Automated Scheduling & Monitoring - Configure suites to run hourly or daily automatically.
  4. πŸ“§ Smart Notifications - Get an email notification when a test suite fails.

πŸŽ‰ That's it! Visit http://localhost:3000 to start creating your first test suite.

πŸ”§ Environment Variables

Create a .env file with the following:

# Required: BrowserUse API Integration
BROWSER_USE_API_KEY=your_browseruse_api_key_here

# Database Configuration
DATABASE_URL=postgresql://postgres:postgres@postgres:5432/qa-use

# Optional: Email Notifications
RESEND_API_KEY=your_resend_api_key_here

# Optional: Inngest Configuration (for production)
INNGEST_SIGNING_KEY=your_inngest_signing_key
INNGEST_BASE_URL=http://inngest:8288

Test Suite Interface

πŸ’‘ How It Works

🎯 Intelligent Test Execution

  1. Define Your Test: Write test steps in natural language

    Steps:
    1. Go to example.com
    2. Click the search button
    3. Type "laptop" in the search field
    4. Press enter and wait for results
    
    Success Criteria:
    The page should show at least 3 laptop search results
    
  2. AI Agent Takes Over: BrowserUse AI agents execute your test

    • Navigate to pages like a human user
    • Interact with elements intelligently
    • Handle unexpected popups and dialogs
    • Adapt to layout changes and dynamic content
  3. Smart Validation: AI evaluates results against your criteria

    • Checks final page state precisely
    • Compares actual vs. expected outcomes
    • Provides detailed pass/fail reports
    • Captures screenshots and recordings

🀝 Contributing

We welcome contributions! This project showcases the full capabilities of BrowserUse and serves as a reference implementation for AI-powered testing platforms.

Development Setup

# Start development environment
docker compose -f docker-compose.dev.yaml up --watch

# Run type checking
pnpm run test:types

πŸ“š Resources & Links

πŸ“„ License

This project is open source and available under the MIT License.