I have quite a lot of small and medium developments in C++/C# (mostly in dotnet) scattered across different repositories, including tools/scripts/configs/docker files
So the idea of creating a single SDK-like repository sounds natural.
But does it really make sense?
There are already countless SDKs, helper libraries, utility packs, boilerplates, and framework-specific toolkits on GitHub. In many cases, whatever I might want to build here are already exists in one form or another. That raises a fair question - Why create another one SDK?
But yes, this personal SDK can be a place where I collect things exactly the way I use them with my preferred abstractions/defaults/structure and with the practical combinations of tools I actually need in real projects
On the other side, we are already living in a time when many programming/infrastructure/ CI/CD /automation/integration tasks can be implemented very quickly by using LLMs, MCP Services and AI agents.
If AI can generate a clean solution in minutes, does it still make sense to maintain this personal SDK repository? Or is it better to generate small focused tools when needed and keep repositories narrower and simpler?
At the moment, I am not sure.
That's why this repository starts as an empty.
Do we actually need this repository now, in the AI era? :)
Maybe yes. Or maybe no.
Let's see!