🔗 Playwright + MCP Integration

This repository demonstrates how to connect Playwright-based Python UI tests to MCP (Model Context Protocol) for intelligent testing flows.

Features

• ✅ Browser automation with Playwright
• 🤖 AI-aware testing using MCP SDK
• 🧩 Configurable YAML for connecting to custom MCP servers
• 💡 Extendable design for adaptive and self-healing test logic

NOTE: pom.xml is for local-use only

Legacy Java Artifacts (pom.xml)

This project is primarily a Python / Playwright / MCP framework. The file pom.xml was retained for local reference/legacy compatibility, but is not committed to this repository.
To keep it locally without committing:

1. Add pom.xml to .gitignore (already included).
2. If pom.xml was previously tracked, run:

   git rm --cached pom.xml
   git commit -m "Remove pom.xml from Git and ignore it"
   
   

---

⚙️ Environment Setup

🪟 Local (Windows)

1. Create and activate virtual environment

   python -m venv .venv
   .\.venv\Scripts\Activate.ps1
   

2. Install dependencies

   python -m pip install --upgrade pip
   python -m pip install -r requirements.txt
   playwright install
   

3. Run tests manually

   pytest -v --maxfail=1 --disable-warnings --html=report.html --self-contained-html
   

Optional: Run manual test flow

Instead of pytest, you can execute the main integration manually:

python main.py

This runs the MCP + Playwright flow directly and prints progress to the console.

GitHub Actions (Ubuntu CI)

This project includes a cross-platform GitHub Actions workflow at:

.github/workflows/python-ci.yml

💡Highlights:

1. Auto-runs on push and pull_request to main

2. Sets up Python 3.11 and caches dependencies

3. Creates a virtual environment automatically

4. Runs Playwright tests and uploads the HTML test report as an artifact

5. Compatible with both Linux (CI) and Windows (local)

🤖 MCP Server Integration

The framework uses the following environment variable:

MCP_SERVER_URL=wss://mcp.openai.com/v1

This enables bidirectional WebSocket communication with the MCP server for AI-assisted actions and validations.

🧪 Test Execution Flow

graph TD
    A[Git Commit / Push] --> B[GitHub Actions Trigger]
    B --> C[Setup Python & Install Dependencies]
    C --> D[Run Playwright + Pytest]
    D --> E[Generate HTML Report]
    E --> F[Upload Artifact to GitHub]

Loading

📊 HTML Report

After each CI run:

1. Report artifact name: pytest-report

2. Accessible in the Actions tab → Artifacts section

3. Local report: report.html created in project root

🚀 Future Enhancements

• Matrix build for Ubuntu + Windows

• CodeQL security scan workflow

• AI-based test case suggestion integration via MCP

• Playwright parallel execution configuration

• Automatic GitHub Pages publishing of HTML report

💡Summary

This framework demonstrates:

• Seamless Playwright + MCP integration

• Unified test setup for both local and CI environments

• Intelligent, maintainable, and extensible Python automation architecture