This is an experimental framework for recording audio that is synchronized to actions on a web page, which can be exported as a web-native presentation.
The framework is intended to decouple audio recording from the generation of synchronization metadata. The pilot "plug-in" uses DZSlides, with an aim to make it easy for anyone to record audio for their DZSlides-based presentation and output the synchronized "movie" as a popcorn-powered presentation. However, similar plug-ins could be written to support the recording of code editing, painting, or anything else web-based.
The recorder uses wami-recorder to record audio via Flash, and delegates the generation of JSON synchronization metadata to the external plug-in. Playback doesn't require Flash and is fully web-native; eventually, even recording will be fully web-native, once browsers start supporting WebRTC.
Currently the project is highly experimental and completely undocumented, however, so play with this at your own risk. :)
I used Improvised with DZSlides to create a narrated slideshow about a CSS selector game that I made.