QA Forum 2025 5 12

🤖 AI + Testing Hackathon Takeaways

Last Wednesday, several backend developers participated in the AI + Testing Hackathon. During the QA Forum, we shared insights on using AI tools—especially Cursor—to improve our testing practices. Here's a summary of what was presented and discussed:

🧪 Demos

1. Martin demonstrated how to use the Cursor AI Agent to create Majority Check integration tests.
2. Li showed how to:
   • Use Cursor's AI conversation feature to analyze and refactor regression and integration test coverage for the Mpay PaymentController.
   • Rule settings and generate rules directly from conversations.

💡 What We Learned

• ✅ AI excels at analyzing and generating documentation.
• 🛠️ Start with simple projects or a limited scope, and iterate with step-by-step prompts.
• 📚 Having sample code and reference documentation is very helpful.
• 🧠 Learned some thinking patterns and tips for creating Cursor rules.
• 🔍 It's essential to carefully review AI-generated code before using it.

🔐 Additional Insights from Anders, Lars, and Victor

• Tool usage of reviewing AI-generated code for correctness and security.
• Highlighted that beyond Cursor, other Agentic AI tools are available:
  • Rider + Windsurf plugin
  • VSCode + Agent Mode with Copilot

🏁 Next Steps

We encourage developers to begin experimenting with AI tools like Cursor to enhance our real-world testing. Start small, review carefully, and share your learnings!