Added and deleted rows now display in a more human-friendly format than raw JSON:
$ csv-diff one.csv two.csv --key=id
1 row changed, 1 row added, 1 row removed
1 row changed
Row 1
age: "4" => "5"
1 row added
id: 3
name: Bailey
age: 1
1 row removed
id: 2
name: Pancakes
age: 2
Example commit here: https://github.com/simonw/sf-tree-history/commit/184efa - compare with https://github.com/simonw/sf-tree-history/commit/dd08cc which demonstrates the previous behaviour.