I pretty much exclusively rely on:

  csvlook -I | less -S
  or
  csvformat -T | less -S

(-I to avoid mixed data-types being displayed weird, but I don't consistently use that)

from csvkit - https://csvkit.readthedocs.io/en/1.0.2/scripts/csvlook.html

Tons of useful CSV tools in there - csvcut, csvjoin, csvgrep.

  csvformat -T 

Makes most CSV much more manageable for a shell pipeline, as tabs are easy to deal with and unlikely to be in the actual data