Finance & Analysis

Verified against ChatGPT · 2026-08-08

Rebuild a personal or team monthly budget from scratch when the old spreadsheet stopped matching reality

Turns a rough list of income and expense line items into a zero-based monthly budget with every dollar assigned a job, plus a named list of assumptions the requester needs to confirm before trusting the numbers.

ChatGPT (GPT-5.1)5 fillable variables

The prompt

Ready to copy — highlighted parts are example details you can swap.

You are helping rebuild a monthly budget from zero because the existing one has drifted so far from actual spending that patching it is no longer worth it. This is a structuring exercise, not financial advice — you are organizing the requester's own numbers into a clearer shape, not telling them what they should be spending or saving.

INCOME SOURCES
Salary $5,400/month after tax, freelance design work averaging $600/month over the last 3 months.

KNOWN RECURRING EXPENSES
Rent $1,800, car payment $310, phone $65, gym $40, streaming subscriptions $38 total.

VARIABLE SPENDING CATEGORIES
Groceries ($400-550), dining out ($150-300), gas ($120-180).

BUDGETING METHOD REQUESTED
Zero-based — every dollar assigned somewhere, nothing left unaccounted for.

TIMEFRAME
Calendar month, starting the 1st, aligned to when rent is due.

RULES
Build the budget so every dollar of listed income is assigned to a category, a savings goal, or an explicit "unassigned" line — never let a difference between income and listed expenses vanish silently. Where a recurring expense could plausibly vary month to month (utilities, variable-rate anything), flag it as an estimate rather than presenting it with the same certainty as a fixed rent or subscription figure. If the requested budgeting method (zero-based, 50/30/20, envelope, etc.) doesn't fit cleanly with the categories given — for instance a 50/30/20 split when no expense has been tagged as a "want" versus a "need" — say so explicitly and propose the smallest change needed to make the categories usable rather than forcing a fit. Do not invent an expense category the requester didn't mention just because it's common (e.g. don't add "entertainment" if it wasn't listed); ask instead. Close with a short list of every number you had to assume or estimate rather than were given directly, so the requester knows exactly what to verify against their bank statement before using this.

OUTPUT FORMAT
1. A table: category | monthly amount | fixed or estimated | % of total income.
2. The unassigned/surplus or shortfall line, called out plainly.
3. Three to five specific follow-up questions that would tighten the accuracy of this budget.
4. The assumptions list described above.

Customize

Optional — swap in your own details for the highlighted parts above.

Why this works

The instruction to assign every dollar to a category or an explicit "unassigned" line closes the single most common failure mode of AI-generated budgets: the model quietly reconciling income and expenses by adjusting a category total until the math looks clean, which produces a budget that balances on paper but no longer reflects what the requester actually listed. Forcing GPT-5.1 to flag estimated versus fixed figures matters because the model has no way to know a utility bill varies seasonally unless told, and left unprompted it will present a single guessed number with the same confidence as a fixed rent figure, which is exactly the kind of false precision that makes a budget look authoritative right up until it's wrong. Refusing to invent unlisted categories addresses a specific pattern where language models pattern-match to "what a typical budget looks like" and pad in categories like entertainment or subscriptions from training-data priors rather than sticking to what the requester actually described — a fabricated category is worse than a missing one because it looks like real input. The closing assumptions list exists because a budget is only as good as the numbers behind it, and naming every place the model had to infer or estimate turns the output into a checklist the requester can verify against a bank statement rather than a black box they have to trust wholesale.

What you get back

Category table: Rent $1,800 (fixed, 30%), Groceries $475 (estimated, midpoint of range), Car payment $310 (fixed)... Unassigned surplus: $312/month. Assumptions: assumed freelance income averages $600 based on a 3-month sample, not guaranteed each month; assumed groceries at range midpoint since no exact figure was given.

Verified against

ChatGPT GPT-5.1 · 2026-08-08

Changelog

  • 2026-08-08 Initial publish, verified against ChatGPT GPT-5.1.

Need this built into your business?

If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.

EXPLORE WHAT SCULT BUILDS
Turn a raw expense export into an anomaly-flagged summary before it goes to whoever approves itStructures a messy expense transaction list into categorized totals and calls out specific line items that look like duplicates, out-of-policy amounts, or unexplained spikes, so the approver isn't the first person to notice a problem.ChatGPT (GPT-5.1)2026-08-09Project cash runway from a bank balance and known inflows/outflows before a hard funding or hiring decisionBuilds a week-by-week or month-by-month cash flow projection from a starting balance and known future transactions, surfacing the exact date the balance would go negative under stated assumptions so a runway decision isn't based on a gut feel.ChatGPT (GPT-5.1)2026-08-09Get a structured first read on a company's financial statements before a meeting where you'll be asked about themSummarizes an income statement, balance sheet, and cash flow statement into the handful of numbers and trends that actually matter for the stated purpose, flagging what needs a follow-up question rather than glossing over gaps.ChatGPT (GPT-5.1)2026-08-10Calculate and interpret a company's key financial ratios against its own history and stated peersComputes liquidity, profitability, and leverage ratios from provided financial figures and explains what each one implies in plain language, without pretending a ratio in isolation proves anything about the company's overall health.ChatGPT (GPT-5.1)2026-08-10
All Finance & Analysis prompts

Check your AI visibility

One URL in, a 0–100 score and the exact fixes out.

RUN THE CHECK

Browse all the tools

15 tools across six categories
13 of them never send your data anywhere

Free · No signup · No trial clock

SEE THE DIRECTORY