CLI: How to ensure it is scriptable

The joystream/cli has good support for interactive style which is very helpful and user friendly for some of the commands that require a lot of input/arguments from the user, prompting user to enter values for fields, checking validity, offering list of choices etc. The output in many cases is presented in a human readable form, sometimes single liner and in many cases multiple tables with various number of columns.

In order for the cli to be script-able there are two key requirements:

1. It should be possible to not require any interactivity, and therefore we must ensure that all commands can be executed by providing all inputs through commandline arguments or by reading a file from stdin. I believe this is already the case for many commands.
2. The output should be optionally formatted in a simple to parse format, so the output can easily be piped to other commands and even the joystream/cli itself. Given the rich output many of the command output, JSON seems like a viable choice.

eg. I want to query MyVideos and get details of the curation status of one particular video. Currently this involves:

cli media:MyVideos
looking in list of videos for entity id of the video of interest
cli content-directory:entity $VIDEO_ID
looking in the properties table for the entity id that represents the curation status:
cli content-directory:entity $CURATION_STATUS_ID

If the MyVideos command outputted json like

[{
  "id": 5,
  "channel": 7
  "title": "MyVideo 1"
},
{
  "id": 6,
  "channel": 7
  "title": "MyVideo 2"
}]

and the content-directory:entity command could read a line from stdin to get the entity id, and output json like:

{
 "id": 5
 "properties": [ 
    a,
    b,
    c, // curation status entity id
 ]
}

we could write a command like:

cli media:MyVideos | jq '.[0] | .id' | cli content-directory:entity | jq '.properties | '.[2]' | cli:entity

┆Issue is synchronized with this Asana task by Unito

Start with the media:MyVideos and content-directory:entity command entry points described in the issue, and trace how they gather input and render output. Define the scope for non-interactive stdin or arguments and machine-readable JSON, then verify the example pipeline works end to end; the issue names no files or tests.