After multiple iterations based on the direct feedback of non-programmer developers, I focused on reducing compilation time, improving failure detection, and providing clearer feedback. The tool reached a point where artists and designers could reliably open the editor with a single action after syncing, without needing to understand or manually trigger recompilation steps.

One key takeaway was how critical fast and explicit feedback is for non-programmer users. Silent failures or long waits significantly reduced trust in the tool early on.

A notable technical trade-off was implementing the tool in PowerShell. While this worked well within a Windows workflow and deepened my understanding of the environment, it limited cross-platform support. Rewriting the tool in Python would improve maintainability, enable iOS/macOS workflows, and allow faster iteration using existing test-driven development practices and cross-platform libraries.