I’ve been building a tool called Wan2.7-Image after getting frustrated with how random most AI image generators feel in real workflows. They can create impressive images, but keeping identity, composition, and style consistent across multiple outputs is still hard. I wanted something more practical for creators and small teams who work from one or more reference images and need cleaner iteration instead of starting from zero each time. If anyone here uses AI image tools, I’d love to know what frustrates you most: randomness, consistency, editing, or control? You can check the project here: htt