Every time you walk away from a Cursor chat for too long and come back to find your context cache has expired, you just paid for that entire context window again. A new open-source extension called Cache Timer aims to fix that by showing you exactly when your cache is about to drop off.
The problem is straightforward. AI coding assistants like Cursor cache your conversation context so follow-up requests are cheaper and faster (the model doesn't have to re-read everything). But these caches have a time-to-live, or TTL - a countdown before they expire. Miss the window, and your next prompt re-processes the full context at full price.
What It Does
The extension sits in your Cursor sidebar and tracks active cache timers across your chats. You can see at a glance which conversations still have warm caches and which are about to expire. The idea is simple: if you know your cache drops in 2 minutes, you can fire off that follow-up question now instead of paying for a full re-read later.
The developer built it after noticing that most unnecessary cache rebuilds on their team came down to two mistakes: forgetting which chat to continue, and forgetting to review AI-generated plans before the cache expired.
Is It Worth Installing?
For solo developers on small projects, the savings are probably negligible. But for teams running Cursor on large codebases where context windows are consistently full, cache misses add up. The extension is free and open-source on GitHub, so the cost of trying it is basically zero.
It's a niche tool solving a niche problem, but it's the kind of small quality-of-life improvement that the AI coding tool ecosystem needs more of - practical fixes for real workflow friction rather than flashy features.