order-status-summary

Purpose

Read an order export CSV and generate a plain-text report that classifies each order into Delivered, Cancelled, Shipped on time, In progress (not yet shipped), Delayed (not yet shipped), or Delayed (past expected delivery).

Inputs

• A CSV path or pasted CSV content containing order records.
• Expected columns: order_id, customer_name, order_date, ship_date, expected_delivery_date, and status.
• No file modifications are performed.

Steps

1. If a CSV path or pasted content hasn't already been supplied (via $ARGUMENTS or earlier in the conversation), ask via AskUserQuestion how to provide it: Provide a file path / Paste the CSV content directly / Use the file already referenced earlier in this conversation (include this option, marked Recommended, only when one genuinely exists). Then request the actual path or content as free text per their choice -- the CSV data itself has no finite menu. If it was already supplied, confirm it looks like a valid CSV reference and proceed.
2. Read the CSV in read-only mode and validate required columns.
3. Map equivalent column names if the user has a slightly different header, and note the mapping used.
4. Parse each row and classify it into one bucket:
   • Delivered: status indicates delivered/completed.
   • Cancelled: status indicates cancelled/refunded.
   • Shipped on time: has ship_date and is within expected delivery timing.
   • In progress (not yet shipped): no ship_date and 3 days or fewer since order_date -- still within normal processing time, not delayed.
   • Delayed (not yet shipped): no ship_date and more than 3 days since order_date.
   • Delayed (past expected delivery): expected delivery date has passed without delivery.
5. Count total orders and bucket totals.
6. Build a delayed-order table sorted by days overdue, with most overdue orders first.
7. Report any skipped or invalid rows and why they were skipped.

Output

• A clear summary report with total orders, counts per bucket, and a flagged delayed-orders table.
• Notes about any missing or malformed data.
• No file writes, no external data transmission.

Guardrails

• Do not invent missing order details.
• Do not modify the CSV.
• If required columns are missing, stop and ask via AskUserQuestion how to proceed: Provide a corrected file (Recommended) / Manually tell me which columns map to the missing ones / Cancel. If mapping manually, ask directly as free text which column corresponds to which.

Notes

• Clickable-question convention: the intake-method choice (Step 1) and the missing-columns recovery choice (Guardrails) use AskUserQuestion. The CSV path/content itself and column-mapping details stay free text -- genuine data, not a finite menu.