SMF·QUICKEN-SIMPLIFI
Can a prompt replace Quicken Simplifi?
Personal finance — budgeting, invoicing and small-business finance
Exhibit tracking slip
Verdict
Simplifi's spending plan is a forecast, not a budget: it takes your recurring income and bills, subtracts what has already gone out, and shows one number for the rest of the month. That arithmetic is an afternoon. The forecast only works if the transactions are there, and they arrive through bank aggregation that a personal build cannot buy. Feed it stale manual exports and the number is confidently wrong, which is worse than having no number.
Exhibit A — The prompt
Received on31.07.2026Build a local spending forecast tool: not a budget, a projection of what is left.
Import: CSV and OFX from files the user downloads from their bank. Saved per-institution column mappings, duplicate detection on a date-amount-description fingerprint, and an import log that can be reversed.
Recurrence detection, which is the foundation everything else stands on. After each import, scan the transaction history for repeating patterns: group by a normalised description (lowercase, strip trailing digits and reference numbers), and within each group look for a consistent interval (weekly, fortnightly, monthly, quarterly) and a stable-ish amount. Surface candidates for the user to confirm, edit or reject — never auto-accept. A confirmed recurring item stores its expected amount, its interval, its next expected date and a tolerance band.
The projection: for the current period, compute expected income still to arrive, expected bills still to be paid, actual spending so far, and from those a single safe-to-spend figure for the rest of the period. Show the arithmetic openly as a waterfall — this number is only trustworthy if the user can see every term in it. Flag any recurring item that is overdue (past its expected date with no matching transaction) or off-band (arrived at an unexpected amount), because those two cases are what make a projection wrong.
Data freshness, which the interface must take seriously: show the date of the most recent transaction in every account, prominently, next to the projection. If any account has not been updated in more than a chosen number of days, mark the projection as stale rather than displaying it as fact.
Also build: category spending against a soft monthly target with a trend line, a rolling 12-month view of every recurring item so subscription creep is visible, and CSV export of everything.
Storage: SQLite with automatic backup before any import.
Out of scope: connecting to any bank or aggregator, moving money, investment tracking, and mobile apps. Say in the README that the projection's accuracy is bounded entirely by how recently the user imported, and that this is the real trade against a connected product.Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 bank aggregation — every figure is only as fresh as your last manual export
- B.2 the mobile app, which is where a spending figure is actually checked
- B.3 merchant names cleaned into something readable
- B.4 the watchlists and alerts that fire when a category runs hot
Prior art
Exhibit C — Why people still pay: financial integrations, compliance, and trust
Because a spending plan is only useful if it is right this morning, and keeping it right means transactions arriving on their own. Six dollars a month for that is not an unreasonable price.
Questions
Can I import my Simplifi history?
Simplifi exports transactions as CSV, which imports cleanly and gives the recurrence detector enough history to work with straight away. Your configured spending plan, watchlists and category customisations do not export and need rebuilding.
How well does recurrence detection actually work?
Well for fixed subscriptions and salary, which is most of what matters. Poorly for variable bills like utilities, and badly for merchants that change their descriptor every month. The confirm-or-reject step exists because the false positives are frequent enough that auto-accepting would poison the forecast.
What does it cost to run?
Nothing beyond the machine it sits on. A SQLite file, no server, no API keys. The cost is your time: a manual export from each account every week or two, without which the projection quietly stops meaning anything.
What is the one thing that does not survive the rebuild?
Freshness. Simplifi's number is right because transactions arrive within a day. Yours is right for as long as your last export was recent, and a spending figure computed from three-week-old data is not a lesser version of the feature — it is a wrong answer presented confidently.
Related tools
Receipt