Rather than a specific skill*, I would instead recommend working on a long-term project of your own design. Build something you find interesting, something so big, and with too many unknowns, that you have no idea how to make it real (yet).

Then start chipping away at it and use the points you get stuck at as places to go learn new things so you upskill. It is this meta-skill* that is among the most important to learn.

Another thing you might do, since you are at uni, is reach out to professors about their research. I did this early in my CS edu and I ended up staying for many years and earning a PhD