Agent reasoning

I am exploring the possibility of enabling reasoning for an agent and was wondering if there is any way to access the generated “reasoning”.