Thread by @HowToAI_

I dropped a full year of bank statements into Claude Opus 4.7 and it sorted everything in 20 minutes.

Income, expenses, recurring subs, the whole picture.

That's easily $11k in accounting work done for the price of a coffee.

Here is the exact prompt I used:

You are acting as my senior forensic accountant and financial analyst. I am uploading 12 months of bank statements (checking, savings, and credit card). Your job is to transform this raw data into a complete, CPA-grade financial picture I could hand to a bookkeeper, lender, or tax preparer.

Work through the following in order, and do not skip steps:

1. Data extraction & integrity check

Parse every transaction across all statements: date, description, amount, account, running balance.

Flag any gaps in statement coverage, duplicate charges, or transactions that look like data-entry/OCR errors.

Reconcile beginning and ending balances for each account, each month.

2. Income analysis

Identify and categorize every deposit: W-2 wages, 1099/contract income, business revenue, transfers between my own accounts (exclude from income), refunds, reimbursements, investment income, gifts, loans, and "other."

Give me monthly income totals, a 12-month total, and the average + median monthly income.

Flag any irregular or one-time deposits I should be aware of.

3. Expense categorizationCategorize every outflow into standard accounting buckets:

Housing, Utilities, Groceries, Dining, Transportation, Insurance, Healthcare, Debt Service, Subscriptions, Entertainment, Travel, Personal Care, Childcare/Education, Business Expenses, Taxes, Charitable, Fees/Interest, Transfers, and Uncategorized. For anything ambiguous, make your best guess and mark it [REVIEW].

4. Recurring charges audit

List every recurring/subscription charge with merchant, cadence (monthly/annual), average amount, and annualized cost.

Flag duplicates, price increases over the year, and anything that looks dormant or forgotten (think: unused streaming, software, gym, app subscriptions).

Call out the top 5 candidates to cancel based on cost vs. likely usage.

5. Cash flow summary

Month-by-month: total income, total expenses, net cash flow, and savings rate.

Full-year totals for the same.

Identify the three highest-spend months and what drove them.

6. Patterns, anomalies & insights

Unusual transactions, potential fraud signals, or merchants that don't match my normal spending profile.

Categories where I'm spending meaningfully above typical benchmarks for my income level.

Three to five specific, dollar-quantified recommendations to improve cash flow over the next 12 months.

Output format:

Start with a one-page executive summary (the numbers that matter most).

Then the detailed breakdowns in clean tables.
End with the recommendations, ranked by annual dollar impact.

Ask me clarifying questions ONLY if something is truly ambiguous and would materially change the analysis. Otherwise, make reasonable assumptions and note them.