As I dive deeper into app development using ShipKitAI, I’ve finally run into the issue of context limits. I’m using Claude Code on the Pro plan, and I’m trying to understand how others handle this. When your context window resets, how do you make sure the model picks up exactly where you left off? Do you create an additional “status” or “progress” markdown document that the code assistant can re-load to re-establish the full implementation state?
Also, I’m considering an additional AI subscription (ChatGPT) since it offers Codex-style capabilities. Has anyone here used it, and how does it compare in practice?