Complete Guide

🧪 TestPilot AI — Complete Wiki

🏠 Home

TestPilot AI is an AI-powered functional testing platform for web, mobile, and desktop applications.

It allows users to:

Write tests in plain English
Record real user interactions
Automatically generate test cases
Run deep site audits

👉 Designed to eliminate scripting and reduce test maintenance.

🚀 Installation Guide

Prerequisites

Node.js (v18+ recommended)
npm
Playwright
(Optional) Appium for mobile testing

Setup

npm install --legacy-peer-deps
npx playwright install chromium

cd client
npm install --legacy-peer-deps
cd ..

npx ts-node server/index.ts

cd client
npm run dev

Open: http://localhost:5173/

▶ Run Your First Test

Enter website URL
Write test in plain English:

verify header is visible, scroll to bottom, take screenshot

Click RUN TEST

✅ Results include screenshots, pass/fail status, timing

🎥 Record & Replay

Record

Start recording
Perform actions in browser
Steps captured automatically

Replay

Execute recorded steps
View live execution

Use Cases

Regression testing
User journey validation

🤖 Smart Test Suggester

Automatically scans website and suggests:

Login tests
Navigation tests
Form validations
CTA interactions

Each suggestion includes:

Priority (High/Medium/Low)
Pre-written test

🔍 Site Auditor

Analyzes:

Performance (FCP, CLS, TBT)
Accessibility
SEO
Broken links
Console errors

Provides score (0–100) per category.

🧠 How It Works

Flow

Components

NLP Parser: Converts plain English to structured steps
Execution Engine: Runs tests via Playwright/Appium
Smart Selector: Finds elements reliably
Reporter: Generates HTML reports

🏗 System Architecture

Key Modules

Client: UI interactions
Server: API orchestration
Engine: Core execution logic
Drivers: Platform-specific automation

🧩 Smart Selector Engine

Uses multiple strategies:

ID
data-testid
aria-label
name
placeholder
text
XPath

Auto-Healing

If selector fails:

DOM scan
Fuzzy matching
Retry execution

📊 Reports Explained

Each test includes:

Step status
Execution time
Screenshots
Error logs

Audit includes:

Performance metrics
SEO checks
Accessibility issues

🧪 Writing Test Cases

Good Example

click login button, verify url contains 'dashboard'

Bad Example

do login stuff

Tips

Be specific
Use clear selectors
Avoid ambiguity

⚙️ Project Structure

server/: backend
client/: frontend
src/engine/: core logic
src/drivers/: automation drivers
reports/: output

🔐 Authentication

Supports Basic Auth:

Username/password input
Secure handling

🐞 Troubleshooting

Playwright Issues

Run:

npx playwright install

Port Issues

Check if 3001 or 5173 already in use

🚧 Roadmap

CI/CD integration
Parallel test execution
AI self-healing improvements
Visual regression testing

⚠ Limitations

Captcha handling limited
Dynamic selectors may fail in edge cases
Mobile setup requires Appium

🤝 Contributing

Fork repo
Create feature branch
Submit PR

❓ FAQ

Is coding required?

No

Does it support mobile?

Yes (via Appium)

Can tests auto-heal?

Yes

🎯 Vision

To build a universal AI testing agent that replaces manual QA and traditional automation.