[Feature] Allow more advanced configuration of output for the `plan` (and maybe `sync`) commands

[b-per]

The following could be useful:

• Allow showing or not the individual jobs changes along with the "summary" table to stdout
• Provide output with advanced formatting to be embedded as PR comment
  • Markdown table of changes
  • Collapsible section for the details of individual jobs changes

Ideally we would get it working with some formatting compatible with GitHub, Gitlab, ADO at least (the ones with native integration in dbt Cloud)

[christineberger]

Would it be too complex to add a --json-format for log output as well? I think having access to something like that would open up the results for a lot more flexibility, like showing it like this:

1. Update job DAILY_FULL_BUILD on dbt project dbt-demo-snowflake in the QA environment

   config	old value	new value
   triggers.schedule	false	true
   job_type	other	scheduled

[b-per]

That's a good idea too!

Maybe as JSONL and not JSON? e.b. each row would be its own JSON
Would this work for you?

[christineberger]

I think JSONL could work! I'm not clear on the process for looping. Is the difference for working with the output as input just looping through the lines of the log (being sure that each line is to_json'd if needed)?

[b-per]

JSONL would allow streaming the different changes, but after a second thought, it might actually be better to also provide a single JSON object at the end of the run. It would be more convenient to programatically consume it.

The format could be something like:

{
  "job_changes": <individual_changes_for_different_jobs>,
  "summary_changes": [
    {
      "action": ....,
       "type": ....,
        ...
     }
   ]
}