Writing good assertions is crucial for robust software, but it’s often a tedious, manual task. What if AI could automate it? Researchers have developed Assertify, a tool that uses large language models (LLMs) to automatically generate assertions for production code. Think of it like a supercharged spellchecker, not just for grammar, but for the logic within your code. Assertify analyzes your code, understands its intended behavior, and then generates assertions that check for unexpected conditions during runtime. This helps developers catch bugs early, improve code comprehension, and ensure their code does what it’s supposed to do. The research explored several prompt engineering techniques to give the LLM the right context, from basic method signatures to sophisticated few-shot learning with similar code examples. The most effective approach involved showing the LLM examples of how assertions are used in similar code, effectively teaching it to reason about code behavior. Results showed that Assertify could generate syntactically correct assertions with high accuracy, exceeding 97% in some cases, and semantically correct assertions with accuracies reaching 83.5%. While the generated assertions aren’t perfect, they demonstrate the potential of LLMs to automate a critical aspect of software development. Future work focuses on refining the accuracy of generated assertions, supporting more programming languages, and exploring open-source LLMs. The potential impact is significant, freeing developers from tedious manual work and enabling them to build more robust, reliable software.