This is a specific example of converting an NDJSON to CSV file, not a generic tool. You'll notice that there are hardcoded headers from the only subfield I was interested in parsing.
Specifically, this example demonstrates transforming an exported PouchDB database (produced by PouchDB Replication Stream) to a CSV file. My docs had variable keys on them, so I hardcoded all of the possible keys to ensure that every entry was written to the CSV (csv-write-stream will assume the first entry has the correct headers and ignore headers on subsequent entries if they differ).
Install.
npm iRun.
node . path/to/input.txt path/to/output.csv