Cursor CLI
Can we have a feature where we can cap the max number of different tokens usage (output, cache read..).
This is useful for automated workflows, so I don’t have to worry about the agent going crazy and consuming a lot of tokens where I don’t except it to do.
As a workaround for now, I process the streamed output using a wrapper and abort the request after exceeding a max length of characters.
Hi Mahmoud,
Thanks for the suggestion.
You’re probably already aware of these two features, but right now, the two closest options to this are carefully setting your on-demand usage limits, and keeping a close look at your usage on https://cursor.com/dashboard/usage, which shows both your included usage as well as your on-demand API usage.
Another helpful tip is to configure Cursor to always show your usage limits, which you can enable in Cursor settings → Agents → Usage Summary and changing it from the Default Auto to Always.
I know it’s not exactly what you’re looking for, but hopefully this can help you keep track of your usage more closely.
Best,
Kevin