Configured scripts without a run section are silently dropped

[grayside]

If you configure a script such as:

scripts:
  ninja:
    description: Ninja vanish

or

scripts:
  ninja:
    run:

the script does not register with rig project.

There should be a verbose or warning level log explaining the incomplete script configuration.

Bonus points if we implement the check for this problem via a rig project config or rig project validate command.

[febbraro]

We validate every config file when we read it, we should just add that validation there, but have a way to determine what level of severity or scrutiny we place on validation (when running a project command we want lots of scrutiny)

[grayside]

Right, I'm suggesting we should have a command dedicated to that, and wondering if setting it up in that way requires changing our approach for further validation rules.

[febbraro]

What is the rationale for a separate command that you have to run, instead of just reporting these problems when the project specific configuration is in play?

[grayside]

Similar to the rationale of docker-compose config -q or composer validate, it means we can test the correctness/diagnose problems with the configuration before committing changes.

Tightening up the validation without regard for a separate command is the priority, but if possible we should let problems trickle up to the caller to determine what/how to report.

[febbraro]

Resolved in #43. If we want a rig project validate or rig project config then lets open another issues specifically for that.