GPT Engineer has introduced several new features to enhance user control and project management, including the ability to select which branch to commit changes to, catering to those preferring a PR-focused workflow or managing experimental work separately. A new Hot Projects tab showcases trending public projects to inspire users, with plans to improve the recommendation algorithm. The platform now detects network errors more effectively, alerting users to API or external service issues. To support larger projects, Retrieval-Augmented Generation (RAG) technology has been integrated, ensuring relevant code is retrieved for each request, improving performance for complex applications. Additionally, the system now automatically fills in placeholders left by language models to ensure complete and functional code.