AGX uses models to generate software from them. This approach is widely spread in other programming languages like Java or C++. Python is suited well too for the model driven approach.
Model-driven engineering (MDE) is a software development methodology which focuses on creating models, or abstractions, more close to some particular domain concepts rather than computing (or algorithmic) concepts. It is meant to increase productivity by maximizing compatibility between systems, simplifying the process of design, and promoting communication between individuals and teams working on the system.
from Wikipedia at 2010-07-20 Model-driven engineering
While AGX aims to be generic in its architecture it currently supports only UML - the Unified Modeling Language - for its source models. It is a graphical language designed to describe software through diagrams. There are several different types of diagrams available, but at the moment AGX uses only class-diagrams.
AGX can be extended by using profiles and matching generators. Each profile defines a set of features and contains a UML profile loadable in any modern UML modeling tool. It also contains AGX generators bound to the scope of the profile responsible for generating Python code, structures or configuration.
Thats completly wrong. UML helps you to structure and document your software, can be a tool to communicate software architecture to non-programmers, helps you to boilerplate the software. Nontheless using a code generator requires the developer to understand what’s generated, otherwise she gets stuck as soon as something gets wrong with the generated output.