This is a topic close to my heart and I am trying my best to create a serverside platform that can only run open source software. My hope is making digital services transparent half the shenanigans will go away. For two reasons

  1. bad actors can't hide B.S.
  2. good actors have something they can point to that has 3rd party validation.

So for digital services where the source code is always available, my runtime dynamically loads the source code indicated in the URL each request. So you are free to go to the URL directly, and you can figure all the routing out in advance of data leaving your device (in theory).

For the sake of moving fast, the 1st supported code repository the runtime can run serverside code from is Observablehq, which has the additional property of being a code IDE which simplifies greatly how you actually upload code to a public location.

The core building block: https://observablehq.com/@endpointservices/serverless-cells

Some blogspam: https://medium.com/nerd-for-tech/towards-a-better-serverless...

I built the IndiaAuth provider using this transparent technique, so you can see exactly how it works.

https://observablehq.com/@endpointservices/auth