algopytest VS beaker

The library we're integrating is algopytest, and Damian is currently hard at work to write documentation and examples for it. On the simulation part I think the simulate endpoint, once released, will be a game changer. There's a lot of things brewing. We will absolutely write up best practices!

You didn't specify what framework you are using. There are different ones besides the native TEAL: pure PyTEAL, Beaker, Reach and Tealish. They abstract state management differently. Take a look if one better suits your style and needs.

MakerX is our technical partner on AlgoKit, and they've been doing a stellar job. Algokit's goal is to make development extremely easy, it was built with modularity in mind because all software (languages included) ages like milk, so whatever language there will be 30 years from now can still leverage our work. We are championing beaker, we're also trying to achieve vanilla python (this is the language) which is a huge engineering effort. Smart contracts programming will never go away, we can however give ever more building blocks that compose the most common parts of smart contracts. We're evaluating formal verification, have a few options (linked in other replies) let us know what you think!

Lastly if you still really want to get into smart contracts you'll want to use a framework, which you can find here (there's a tutorial on the algo dev docs too I think): https://github.com/algorand-devrel/beaker