You're generally correct that we keep on top of critical security updates for all components of our infrastructure, including the python runtime that python App Engine uses, and python libraries.
However, some aspects of security are up to you as a developer: for example ensuring that your API only processes requests from valid clients, ensuring that users can only access the data on your app that is meant to be visible to them, ensuring that you don't expose yourself to SQL injection if you use a remote MySQL box for data storage, etc.
Using
Endpoints for your API, you can
implement client auth, so there are even aspects of the platform that help you do these things in a way that respects best-practices. There's even a beta
Security Scanner which is very good.
In general, programming any web app, these issues tend to come up. You'll also have to implement protections against XSS and XSRF, although not from scratch, as there are many great libraries and frameworks out there.
Does this help answer your question? Do you have any more specific questions about security?