Gross spending by type (excluding credit notes)

# Instructions — Gross spend by type (with creditNote exclusion)

## ⚠️ Critical rule

`spendesk_analyze_spend` includes `creditNote` by default. Their amount is **negative** — they reduce `summary.totalAmount`.

On an observed demo account: `creditNote` accounted for **-6.5% of the total** (-EUR 58,950 out of EUR 910,699).
Same for `cardRefund`: negative amounts (-0.1%).

**For gross spend**: exclude `creditNote` and `cardRefund` via the `payableTypes` filter.
**For net spend**: include them (this is the default behavior).

## Step 1 — Identify the company and period

Call `list_companies` to get the `companyId`. Define `fromDate` and `toDate` (YYYY-MM-DD, max 366 days).

## Step 2a — GROSS spend (excluding credit notes)

Call `spendesk_analyze_spend` with:
```json
{
  "groupBy": ["payableType"],
  "fromDate": "...",
  "toDate": "...",
  "filters": {
    "payableTypes": ["invoicePurchase","mileageAllowance","perDiemAllowance",
                     "singleCardPurchase","physicalCardPurchase",
                     "multiUseCardPurchase","subscription","expenseClaim"]
  }
}
```
→ `creditNote` and `cardRefund` excluded → `summary.totalAmount` = gross spend

## Step 2b — NET spend (with credit notes)

Call `spendesk_analyze_spend` without the `payableTypes` filter.
→ `summary.totalAmount` includes credit notes (net accounting value)

## Step 3 — Compare and present

Present both totals:
- Gross spend: X EUR
- Credit notes and refunds: -Y EUR  
- Net spend: X - Y EUR

By type:
| Type | Amount | Share |
|------|---------|------|
| invoicePurchase | ... | 93.1% |
| subscription | ... | 9.7% |
| ... | | |
| creditNote | -... | -6.5% |

## Example prompt

- "What is our gross spend in June 2026?"
- "Show me the total spend by type this month, excluding credit notes"