There's also NAND to Tetris [1], csprimer [2], and/or lists like these [3] or codecrafters [4].
[0] https://third-bit.com/sdxpy/
[1] https://www.amazon.com/Elements-Computing-Systems-second-Pri...
[2] https://csprimer.com/courses/
[3] https://www.andreinc.net/2024/03/28/programming-projects-ide...
They both come from the Racket school of pedagogy and focus on systematic program design [2]. The first resource is free and excellent, but the second goes into more detail.
[0] https://htdp.org/2024-11-6/Book/index.html
[1] https://www.amazon.com/stores/author/B09QMPRYYM/allbooks?ccs...
[2] https://felleisen.org/matthias/Thoughts/The_Design_Recipe.ht...
What worked for me was writing an image editor. Lots of different filters and effects I could add to it.
Then games; if you do that, start with minesweeper or chess (without AI), add features and complexity to do whatever seems fun. You don't need to bother with a fancy game engine at this level, vanilla JavaScript or python or whatever else you're using is good enough, everything is good enough.
If you love sounds and how they work, make a sound processor. Music, procedural music composition.
Main thing is just keep yourself excited for the next step; what that means for you is something only you can answer.
If chess appeals, how about chess variants, perhaps one like Penultima where your implementation could read a custom secret rule for a piece and allow/forbid according to that rule?
(This is also one of the projects on my to-do list).
Games in general can become arbitrarily complex: A tower defence game can expand to become all of Starcraft, or a Settlers of Catan clone could expand to become Civilization.
But yeah, focus on something you'd enjoy either using or playing, depending on if it's a game or a tool.
For programming, learn a new thing to do with your language (network programming, say). Or learn a second language.
For software engineering... that winds up being all about the accurate communication of technical information, so work on reading and writing technical documentation. Also more architecture-level issues rather than fits-in-a-few-lines-of-code kind of problems.