AI-powered dynamic screenshot generation and annotation

I would like to request a feature where AI image generation capabilities could extend to creating screenshots dynamically and adding annotations, such as arrows, on the fly. This would greatly enhance the ability to provide up-to-date and context-specific visual aids.