To fix my own workflow, I built cnsplots. It’s a lightweight wrapper that generates publication-ready figures (like those in Cell, Nature, and Science) with zero configuration, letting you focus on the data instead of fighting plotting aesthetics.
I'd love for anyone who regularly builds Python data visualizations to try it out. Open to any feedback or hearing how you handle standardizing your own plots!