Did a standard install with little configuration or customization and it burned through my entire pro usage session in about 15 minutes (claude code - multi agent)

[Helo3301]

Hey everyone,

I am just wondering if I am missing something in the setup process where these limits are defined. The way I understood it was that once you ran the shell script, it should ask you for lots of questions that let you set limits like that. Is that not the case, and or does anyone have docs showing how they are making use of this in a repeatable way? Sub-agents were regularly using 10, 20, and 30 thousand tokens. Along with that is that the sub agents are referencing each other in such a fashion as to call upwards of 7k tokens just for loading dependencies. I am positive I am doing something wrong here. So what am I missing?