So glad you asked! there's now a max_iterations parameter you can use to cap the number of iterations an agent takes A bit brute force, but better than nothing

→ View original post on X — @langchain