/money-mirror

---
name: money-mirror
description: >-
  Find the emotional patterns behind someone's spending, saving or
  investing decisions: reads a bank export, a spending list or a short
  diary of money moments, groups them by the feeling that came before
  (stress, boredom, reward, guilt, comparison, fear of missing out), names
  the behavioural bias at work in plain words, and suggests one small
  experiment for the next 14 days. Use this skill whenever the user says
  "where does my money go", "I always overspend when", "why can't I
  save", "I feel guilty about money", "emotional spending", "I check my
  investments too often", or shares transactions and asks what they reveal.
  Do NOT use as therapy or diagnosis, for gambling or debt crises (point to
  professional help), or to shame anyone about how they spend.
---

# Money mirror

Most money advice assumes we decide with spreadsheets. We decide with
feelings and explain it with spreadsheets afterwards. A €40 order at 11pm
after a hard day is not a budgeting error. It is a mood with a card number.
Telling that person to "make a budget" fails because the budget never
meets the feeling.

Behavioural finance has names for the patterns: loss aversion, present
bias, mental accounting, social comparison, the ostrich effect. The names
matter less than seeing your own pattern in your own numbers. This skill
holds up the mirror, kindly, and then suggests one experiment small enough
to actually try.

## Workflow

### 1. Collect the moments

Accept any of:
- a bank or card export (date, amount, merchant),
- a typed list of recent purchases or money decisions,
- a short diary: "what I bought, when, how I felt just before".

For an export, ask the user to mark 10 to 20 transactions they remember
well and add one word for the feeling before each. Their words beat any
guess. Remove card numbers, IBANs and other people's names before reading
(see redact-first).

### 2. Find the patterns

Group the moments by:
- **When:** time of day, day of week, after payday, after work.
- **Feeling before:** stressed, tired, bored, celebrating, lonely, anxious,
  compared to someone.
- **Where:** app, shop, website, social media link.
- **Size:** many small vs few large.

Look for clusters of three or more: "7 of 12 food deliveries were after
9pm on days marked 'tired'".

### 3. Name the bias, plainly

For each cluster, one line:

| Pattern | Bias, in plain words |
|---|---|
| Spending more right after payday | Present bias: money now feels bigger than money later |
| "It was on sale, so I saved €30" | Mental accounting: the €70 spent counts as €30 earned |
| Not opening the investment app when markets fall | Ostrich effect: we avoid information that might hurt |
| Checking the app 10 times a day when markets rise | Attention as reward: the feeling, not the information |
| Buying after seeing friends' posts | Social comparison: the reference point is someone else |

### 4. One experiment

Suggest ONE change for 14 days, tied to the strongest pattern, designed to
add friction to the feeling, not willpower:

- Delete saved card details from the delivery app; log in only by hand.
- A 24-hour wait list for anything over €50 bought after 9pm.
- Check investments once a week, on Sunday, with a coffee.

Say what to measure and when to look again.

### 5. Report

```
## Money mirror, [period]

**The pattern that costs most:** [cluster, number, money]

**What I see** (3 patterns max, each with its bias in plain words)

**What's working:** [at least one thing they do well, with evidence]

**Your 14-day experiment:** [one change, what to count, review date]
```

## Rules

**Curious, never judging.** No "should", no "bad", no "wasteful". The words
are "pattern", "when", "what happened before".

**Their feelings, their words.** Never assign an emotion the user did not
report. Ask.

**One experiment.** Five changes at once is a diet that ends on day four.

**Know the limit.** If the patterns suggest compulsive spending, gambling,
debt that cannot be paid, or distress, say so gently and suggest a
professional (a debt adviser, a counsellor, a support line).

## Example

A user shares 3 months of card data and marks 18 purchases. Two clusters
stand out: 9 online clothing orders, 8 of them on Sunday evenings marked
"anxious about Monday", averaging €62; and investment-app logins jumping on
days the market rose. The mirror names "mood repair" and "attention as
reward", notes that their savings transfer happens on payday automatically
(working well), and suggests one experiment: no shopping apps on Sunday
after 6pm for two weeks, replaced by planning one nice thing for Monday.