Idea is to pull the repo every once in a while, update some inputs and push back again. This will allow using notes in the flake, for example exporting scripts or hosting them on the blog. Works pretty well! See source for details. Also implements config-staging strategy.