Streams the project's tickets as a CSV download (tickets.csv, text/csv), with no row cap. For a JSON export use GET /tickets/export instead.

Rows: type narrows the export to one type or a comma-separated list such as BUG,INQUIRY,BOT (BOT is the Kai conversations lane). Without type, tickets of every type are exported, including types whose board was deleted and survey responses. Archived tickets (ignoreArchived is accepted and ignored), spam and merged duplicates are included. startDate/endDate (ISO 8601, inclusive) filter on createdAt; to filter on the update time use updatedAt>=2026-10-01T00:00:00Z&updatedAt<=2026-10-08T00:00:00Z (UTC with a Z suffix: a +hh:mm offset is decoded as a space and matches nothing). The usual ticket filters apply; sort, skip and limit are ignored and rows come in _id order.

Columns: pass projection[] to pick fields (see GET /tickets/export-fields). Without it, the columns are the listed types' export fields, or the BUG board's fields when type is omitted.

Exports ticket records only, not their conversations. Retention limits which tickets still exist — see GET /tickets. If the export fails after the download has started, the connection is aborted rather than ended, so a download that completes without a transfer error contains every matching row.