# Spending Analysis

## Scope and Period

Explain spending and cash flow only for the period and scope actually supplied. Never extend a conclusion beyond the period given — a current-month figure is not evidence about last month or the year.

## Prepared Totals, Rankings, and Comparisons Are Authoritative

If FinAnts supplied a total, a category ranking, a comparison between periods, or a share-of-total percentage, treat it as exact and final. Never recompute it by summing individual line items yourself, never reorder a supplied ranking, and never produce an alternative figure. If your own reading of visible line items seems to disagree with a supplied total, use the supplied total — the visible items may be a partial, recent subset, not the complete period.

## Suggesting a Review or Pause

You may suggest reviewing a category, confirming an unusual pattern, or pausing a specific discretionary expense **only when the supplied data actually supports it** (e.g., a category total or an increase was actually supplied). Never promise a specific saving amount from a suggested change unless FinAnts supplied that impact figure itself.

## Deficit Signal

A negative net cash flow or a spending pattern that exceeds income is the highest-priority diagnostic signal — see [SKILL.md](../SKILL.md) §J. Name it plainly and directly when supplied data shows it, before discussing any lower-priority topic (a goal, an optimization) in the same answer.

## Interpreting a Category, Carefully

Describe what the data shows (amounts, categories, dates, comparisons). Do not speculate about the *nature* of a category beyond what the data supports — for example, do not assume a "Vivienda" expense is a mortgage versus rent, or that high spending in one category signals a specific behavior, unless the data or an explicit rule confirms it. If you offer an interpretation, present it explicitly as one, never as a verified fact.