Tag a Batch of Tickets Into Your Taxonomy as CSV

You are a careful ticket classifier. Apply our existing tag taxonomy to a batch of support tickets and return clean data I can analyse in a spreadsheet.

Our taxonomy, with tag definitions and rules: {{TAXONOMY}}
The tickets, one per line, each with an ID and the subject or text (names, emails, phone numbers, addresses and payment details removed): {{TICKET_BATCH}}

Return one CSV row per ticket with these columns: ticket_id, primary_tag, secondary_tag (or blank), sentiment (positive, neutral, negative), urgency (low, normal, high), confidence (high, medium, low), and evidence (the shortest quote from the ticket that justifies the primary tag).

After the CSV:
1. List the ticket IDs with low confidence, and what made each one ambiguous.
2. List tickets that don't fit any tag, and say whether a new tag is needed.
3. Give the count per primary tag, computed from your rows above.

Rules: use only the tags in our taxonomy, spelled exactly. Don't infer urgency from customer value or anything outside the ticket text. Never fill a field by guessing; write "unclear" instead. Keep each evidence quote under 15 words and copy it exactly. Return every ticket I gave you, in the same order, with no rows merged or skipped.