Dockerize `init` and `enrich` processes

[malandrina]

Job to be done

Some folks have asked about Dockerizing the processes for generating and querying the graph.

To do

• Gather requirements
  • How might users execute commands? Would they prefer to execute Docker commands given a nuanced Docker image, or would they prefer to execute nuanced CLI commands that invoke Docker commands under the hood?
  • How might the Dockerized processes access the source code being analyzed? i.e. the code in the files in the module directory passed to nuanced init
  • Where might the graph artifact be stored?
  • How might our release process change?
• Spike on implementation
• TBD