Create foundation for project pipeline

[Boris Kuznetsov]

We'll have a crawler job which will gather projects repository links
across the web and stash them to the database as a queue. Then we need
to handle this queue.

We should have a rake task, which on invocation will run in a
infinitive loop and run pipeline for each new enqueued link
(projects table).

What needs to be done:

• create a rake task
• create a pipeline service which rake task will evoke
• pipeline service will look for the next link
• will call a method to clone git project locally (mocked)
• will call a method/module which will scan project for improvements
• improvement scanner will apply improvement processors each by each
  (define sample processor as well)
• finalize action which will check if something changed
• will call an internal PR creation service if improvements detected