The concepts/reg-ex redirect is how I was imagining the autonomous agent automation. It does need to be fleshed out a bit more, but essentially, it was an easy way to let a process to know what's happening in another process so it could react appropriately. The e2e tests are a good use-case, but not something I've looked into yet (in theory, the framework should hold up for it)

Original intent was just a hobby/side project. As it grew, I thought maybe posting about it would be alright in case people find the idea interesting.