In the naive version, there would be a single state and no triggered responses.
Would this be as straightforward as defining variables in the initial state, then referencing them within various modules in Kyma sounds, so that upon compilation (or runtime) the data defined in the Tool may be accessed?
For a more elaborate version, if a data set defined within a Tool state can be accessed during runtime, I could see defining multiple states, one for each desired automorphism of the data set. This way perhaps, a single variable accessing the data set might be continuously referenced as an EventValue from a running Kyma sound, while triggers sent to the Tool change its state in real time, to transform the data set according to some group operation.
Relating to the above, is a variable defined within a Tool state (or global state) accessed from a Kyma sound during compilation time, runtime, both, or either (depending on how or where it's defined)?