This parse_transform makes your code self-descriptive. When you add @doc annotation for module or function, it goes to EDoc's output. With autohelp you can view it using Module:help functions:

• Module:help() shows module description and list of functions with names of arguments
• Module:help(FunctionName) shows brief description for functions with given name and all arities
• Module:help(FunctionName, Arity) shows full function description

All output goes to stdout (it is designed to look well in shell), return values are ok.

Just add compile option to your code:

-compile({parse_transform, autohelp}).

After that functions help/0, help/1, help/2 will be added to your module.