Last week I thought: if I use Node-RED/n8n for managing Jira tickets via API, it will be almost the same as my idea of a PN editor. You can't call it process management since state is somewhat hidden: it models only data flow, not state. But this works; I can work with Jira, and now I will try to create some processes outside of Jira to test it out, like an intake process.
I still want to finish the PN task manager, though. Yesterday I spent some time trying to understand where the complexity is coming from and how I can reduce it so I can actually move on. Nothing is clear yet. I read through some colored Petri nets articles and jotted down what might be useful, but I still need more investigation. I don't need simulation and verification, so I can simplify parts of the framework but still keep it useful.