Split trace implementation in two #123
Description
One implementation is intended to be used on fetched repositories, and will always write DataFrames to disk
The other one is intended for inline code usage, and will return built DataFrames to the callee