Enabling Testability: Key Step in Automating Automation

Abstract

In the modern software development lifecycle, automation has become essential for achieving speed, accuracy, and scalability in testing processes. However, a critical step often overlooked is ensuring the system's testability early in its development. This paper introduces the concept of the Testability Analyser, a tool that evaluates software systems for their testability. By leveraging AI technologies and integrating with design tools such as AWS documentation, Draw.io, PlantUML, and Creately, the Testability Analyser facilitates early testability evaluations, optimizing systems for automated testing. The paper discusses the importance of testability, the role of AI systems in understanding complex software architectures, and key items to verify testability in software architecture before and after the advent of Large Language Models (LLMs) and Model Context Protocol (MCP).