🚀 TestPilot.AI

Automated Testing Platform powered by GitHub Copilot

Analyze web applications and generate comprehensive Playwright test suites in seconds.

---

📖 Overview

TestPilot.AI is an AI-driven platform designed to accelerate quality assurance. By leveraging the GitHub Copilot SDK, it crawls target URLs, identifies critical user flows, and automatically generates high-quality Playwright test strategies and code artifacts.

It's not just a code generator; it's an intelligent QA companion that understands site structures and produces valid Page Object Models (POMs) and test specs tailored to the application's unique ecosystem.

✨ Key Features

Feature	Description
🔍 URL Analysis	GitHub Copilot Client extracts title, description, internal links, accessibility tree and form structures.
🤖 AI-Powered Artifacts	Generates detailed test strategies and Playwright code using GitHub Copilot SDK.
⚒️ Smart Validation	Automatically validates generated code for syntax and type errors via the TypeScript Compiler API.
📦 ZIP Packaging	Download the entire Playwright test suite, including package.json and POMs, in a single ZIP file.
🌓 Modern Dashboard	Advanced UI with Strategy View, Code Viewer (with syntax highlighting), and Dark Mode support.
📊 Real-time Progress	Visual feedback during the analysis process with sequential state indicators using GitHub Copilot SDK.

🛠️ Tech Stack

• Frontend: Next.js 16 (App Router), React 19
• AI Engine: GitHub Copilot SDK
• Styling: Tailwind CSS 4, Radix UI, Lucide React
• Parsing: Cheerio
• Utilities: JSZip, React Markdown
• Validation: TypeScript Compiler API
• Testing: Vitest

📂 Project Structure

src/
├── app/                  # Next.js App Router (pages & API)
│   ├── api/agent/        # Core agent logic and backend endpoints
│   └── globals.css       # Global styles (Tailwind 4)
├── components/           # UI Component Library
│   ├── dashboard/        # Analysis results, CodeViewer, StrategyView
│   ├── landing/          # Hero and landing page sections
│   ├── layout/           # Header, Footer, and navigation
│   └── ui/               # Reusable Radix/Shadcn primitives
├── lib/                  # Business logic & utilities
│   ├── mdParser.ts       # Markdown parsing for AI responses
│   ├── parser.ts         # Web crawler and scraper logic
│   ├── validator.ts      # TypeScript syntax validation
│   └── metrics.ts        # Performance and usage tracking
└── docs/                 # Project documentation and roadmap

🚀 Getting Started

Prerequisites

• Node.js: ^20.x or later.
• Package Manager: pnpm (recommended).
• GitHub Copilot: Access to GitHub Copilot (the agent uses the SDK which requires authentication).

Installation

1. Clone the repository:

   git clone https://github.com/fugazi/testpilot-ai.git
   cd testpilot-ai

2. Install dependencies:

   pnpm install

3. Install the GitHub Copilot CLI and the Copilot extension: Installing the GitHub Copilot CLI is necessary to authenticate your Copilot session.

   # Install the GitHub Copilot CLI
   npm install -g @github/copilot-cli
   
   # Install the Copilot extension
   npm install -g @github/copilot

4. Authenticate with GitHub Copilot (Mandatory): An active GitHub Copilot subscription is required to use the GitHub Copilot SDK.

   copilot auth login

Running the Project

# Start development server
pnpm dev

# Build for production
pnpm build

# Run unit tests
pnpm test

# Run linting
pnpm lint

⚙️ How it Works

1. Input: User provides a URL (e.g., https://example.com).
2. Crawl: The system performs a server-side fetch and analyzes the HTML to understand forms, buttons, and links.
3. Agent: The data is sent to the GitHub Copilot Agent with a specialized QA Prompt.
4. Validation: The AI's response is parsed, and any code blocks are validated for syntax correctness.
5. Output: The user receives a structured Test Strategy and a set of Playwright files ready to use.

🗺️ Roadmap & Phases

Check the current progress in our Plan Phases document:

• ✅ Phase 1: MVP Core & Basic Agent integration.
• 🚧 Phase 2: UI/UX Overhaul & Artifact Validation (Current).
• 📅 Phase 3: Playwright MCP Integration for dynamic exploration.

🏠 Developer

• Name: Douglas Urrea Ocampo
• Job: SDET - Software Developer Engineer in Test
• Country: Colombia
• City: Medellin
• E-mail: douglas@douglasfugazi.co
• LinkedIn: https://www.linkedin.com/in/douglasfugazi
• Contact: https://douglasfugazi.co

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Made with ❤️ by Douglas Urrea Ocampo for the QA Community