But as soon as I think about implementation I realize how complicated it is to make such an intelligent system that would understand intent and based on intent adapt the action. In theory, it all can be manually connected, but then it would require a complicated setup.
Of course, it's only from an implementation perspective. From the UX it can be trivial.
Your idea is very good and you can even monetize your app by selling ad space when users choose "boredom" then you can recommend them sponsored apps and games.
This idea would be also interesting on PC, when users lock their screens and then come back to do something. Maybe it can even be part of some diary/note taking app where when you unlock your screen note window pops up and asks you "What are you planning to do now?".