Anyone else feel like autonomous AI coding just overcomplicate basic fixes?
by•
been testing multi-agent setups with tools like cursor and claudecode recently. they were great early on, but giving them more autonomy seems to backfire.
instead of sticking to the main task, they end up hunting down tiny bugs or adding unnecessary code tweaks that just drain hours in loops.
is this just a prompt control issue or are agents hitting a limit with complex context?
13 views
Replies
Recently discussed this same issue, and I think more developers are moving to implementing more control and boundaries into their workflows. More agents create more search space, thus expanding the task. Not completely sure, but I would think complex content would add even more search space, depending on tool calls and other changes within the code. Try implementing more constraints for your agents, that will help them stay on task.
I agree with @jason013 . Unfortunately, they are designed to burn tokens so they will " hunt tiny bugs or adding unnecessary code tweaks that just drain hours in loops." for the sake of it.