A product description says blue, large. If an assistant simply joins values with commas, that description becomes two columns. The file still looks tidy, but every field after it may move to the wrong place.
Agree on the format first
Specify the receiving application, column names, delimiter and encoding. For a comma-delimited UTF-8 file with the columns id,description, this example is one record with two fields: 7,"blue, large". The quotes keep the comma inside the description.
RFC 4180 documents common CSV conventions rather than every application's import behaviour. Fields containing commas, double quotes or line breaks are enclosed in double quotes. A quote inside a quoted field is doubled: 8,"The ""large"" size" represents the description The "large" size.
Use plain ASCII double quotes for CSV syntax. Typographic quotation marks copied from prose do not perform the same job. Do not replace commas globally: some are separators and others are data.
Import a small, varied sample
Include a plain description, the comma example, the embedded-quote example and a description containing a line break. Read the sample with the destination application's CSV importer or a CSV parser. Compare the actual cells, not just the text layout. One record with a line break can occupy several physical file lines.
After the sample works, check record counts, column counts and values in the full export too. A successful small test does not validate unseen rows. Preserve the original list and the import settings. Ask for the actual data file; a table drawn in the chat is not the same deliverable.