pipeline/node inputs same as raw data and make the associated scripts not part of the project source so they are stored in the same level as the data?