When creating a new piece of code, I’d recommend starting with having it split up in quite a lot of code blocks, all covering just a few or even just a single step of the process.
After you get that functional, you can always “summarise” all that in a single code block after. As you can see the result of each individual line of code then, it becomes a lot easier to figure out where the source of a problem would be.
Hi @PauLtus, thanks for replying!
Yeah I usually start with placing individual nodes to the graph and then once they seem to function properly I convert them to code blocks for better legibility.
Although in this case I thought it would be better to use textural scripting for the second part of the graph for desired execution flow. But again, this was before I understood how transaction nodes work.
Alas, lesson learnt! 
It seems you already got the right workflow then. 
Even if I’m pretty confident I can do it one go I tend not to, when something goes wrong and everything is already tightly coded it can take quite some time dissecting to find out what exactly went wrong.
