Very simplistic view on the problem domain IMHO. Yah sure we can add a bunch of functions... ok. But how about snapshotting (or at least work with git), sandboxing both process and network level, prompt engineering, detect when stuck, model switching with parallel solvers for better solutions. These are the kind of things that make coding agents reliable - not function declarations.

It will be included as part of the third instalment. I write these coding agents for a living. Need to start with the basics as the basics is what people need to know to be able to automate functions at their employer, which may not be coding agents. This workshop was delivered at a data engineering conference, for example.