Skip to main content
skill CX Codex

spending-review — Claude Code skill: categorized spending breakdown

This skill groups your spending by category for a period, compares it to the prior month and flags where costs jumped. Useful if you track money in Wilson or bank CSVs.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add openaccountant/skills --skill spending-review -a claude-code
Open on GitHub

What spending-review does

You get a table: category, this month, last month, change in dollars, change in percent, trend arrow. Under it sit the top 3 categories by absolute spend and the top 3 by percentage increase, plus a short written summary of the pattern.

Two Wilson tools do the work. spending_summary returns totals grouped by category for a date range, transaction_search pulls individual transactions. The skill runs spending_summary for the current and previous month, computes the difference, and flags any category up more than 20% or more than $100. Each flagged category gets a transaction_search call with category: "<flagged_category>" and months: 1.

Category accuracy rides on your bank's auto-categorization or Wilson's rules, so run categorize first when many transactions arrive uncategorized. One-off purchases — an appliance, a trip, a medical bill — skew month-over-month numbers, so a spike may be an outlier rather than a trend. The skill suggests a 3-month rolling average as a steadier baseline.

What it helps with

  • break monthly spending into categories
  • compare this month's costs against last month
  • spot categories where spending jumped
  • drill into the transactions behind a pricey category
  • get a budget summary without building spreadsheets by hand

Example prompts for the agent

  • Break down my spending this month by category and compare it to last month
  • Which categories did I spend noticeably more on than last month?
  • Show me the transactions that pushed up my Groceries category

How to install spending-review

1Skills CLI (Vercel)

npx skills add openaccountant/skills --skill spending-review -a claude-code

2Manually: put the files in place

~/.claude/skills/spending-review/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/spending-review/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about spending-review

How do I install spending-review in Claude?

Run npx skills add openaccountant/skills --skill spending-review -a claude-code. The skill lands in ~/.claude/skills/spending-review/SKILL.md.

What does spending-review do?

It produces a categorized breakdown of spending for a period, compares it against the previous month, and flags categories where spending rose by more than 20% or more than $100.

Do I need Wilson to use this skill?

No. Without Wilson the skill walks through a manual route: export two months of transactions as CSV, combine them in Google Sheets, and build a pivot table with change calculations.

Source text

SKILL.md · MIT
---
name: spending-review
description: >
  Generate a categorized spending breakdown with month-over-month trends.
---

# Spending Review

## Overview
Produces a categorized breakdown of your spending for a given period, compares it against prior months to show trends, and highlights categories where spending increased or decreased significantly.

## Wilson Tools Used
- `spending_summary` — get total spending grouped by category for any date range
- `transaction_search` — drill into specific categories for transaction-level detail

## Workflow
1. Run `spending_summary` for the current month to get category-level totals.
2. Run `spending_summary` for the previous month to establish a comparison baseline.
3. Calculate the dollar change and percentage change for each category between the two months.
4. Flag any category where spending increased by more than 20% or more than $100.
5. For each flagged category, run `transaction_search` with `category: "<flagged_category>"` and `months: 1` to pull the individual transactions driving the increase.
6. Present results as a summary table with columns: Category, This Month, Last Month, Change ($), Change (%), Trend Arrow.
7. Below the table, list the top 3 categories by absolute spending and the top 3 by percentage increase.
8. Provide a one-paragraph narrative summary describing the overall spending pattern and notable changes.

## Without Wilson
1. Export your last two months of transactions as CSV files from your bank (most banks: Account > Statements > Download > CSV format).
2. Open both CSVs in Google Sheets. Combine them into one sheet with a "Month" column added.
3. If your bank doesn't categorize transactions, manually add a "Category" column. Common categories: Groceries, Dining, Transportation, Utilities, Entertainment, Shopping, Health, Subscriptions.
4. Create a pivot table: Rows = Category, Columns = Month, Values = SUM of Amount.
5. Add a calculated column for change: `=B2-C2` (this month minus last month).
6. Add a percentage change column: `=IF(C2<>0, (B2-C2)/ABS(C2)*100, "New")`.
7. Conditional format the change column: red for increases, green for decreases (since spending is negative, reverse the logic or use absolute values).
8. Sort by absolute change descending to see your biggest movers.
9. For a quick visual, insert a bar chart from the pivot table showing this month vs. last month by category.

## Important Notes
- Category accuracy depends on your bank's auto-categorization or Wilson's categorization rules. Run the `categorize` tool first if many transactions are uncategorized.
- One-time large purchases (appliances, travel, medical) can skew month-over-month comparisons. Consider whether a spike is a true trend or an outlier.
- For a more meaningful view, compare against a 3-month rolling average rather than just the prior month.

© openaccountant · MIT · openaccountant/skills

Similar skills

Everything for Finance and business →

From the same set28

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor