SMF·POCKETSMITH
Can a prompt replace PocketSmith?
Personal finance — budgeting, invoicing and small-business finance
Exhibit tracking slip
Verdict
Projecting a future account balance by extrapolating recurring transactions forward on a calendar — Pocketsmith's actual signature feature — is a real weekend build once your recurring transactions are correctly identified. What doesn't survive: Pocketsmith's automatic recurring-transaction detection and its polished multi-account forecast blending.
Exhibit A — The prompt
Received on31.07.2026Build a personal finance ledger with a forward-looking balance forecast — Pocketsmith's actual signature feature, distinct from this catalogue's other personal-finance entries which focus on categorization or multi-currency. Use Next.js/TypeScript with SQLite, the same CSV import approach as this catalogue's Tiller entry. Let the user tag any transaction as recurring, with a frequency (weekly, monthly, custom interval) and next-expected date and amount. Build a forecast view: starting from today's actual balance, simulate each recurring transaction forward day by day for the next 90 days, producing a projected balance line chart that shows both the confirmed actual balance (past) and the projected balance (future) on one continuous timeline. Let the user adjust a recurring transaction's amount or skip a single occurrence without deleting the rule, to handle real-world variance. As a nice-to-have, flag transactions that look recurring based on a simple heuristic — same payee and similar amount appearing three or more times at roughly regular intervals — as suggestions the user confirms, rather than claiming full automatic detection. Do not build multi-account forecast blending, live bank-feed connections, or what-if scenario comparison — those are out of scope. Requires hosting and a database; no external API key needed.
Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 automatic recurring-transaction detection
- B.2 polished multi-account forecast blending
- B.3 live bank-feed connections
- B.4 what-if scenario comparison tools
Prior art
Exhibit C — Why people still pay: financial integrations, compliance, and trust
Projecting forward from known recurring transactions is simple math; automatically detecting which of hundreds of transactions are actually recurring, without the user tagging each one, is a real pattern-recognition problem.
Questions
Does it automatically find my recurring bills, or do I have to tag them?
It suggests candidates using a simple heuristic — same payee and similar amount appearing several times at regular intervals — but you confirm each one. Pocketsmith's fully automatic detection is a harder pattern-recognition problem this build approximates rather than solves.
How far into the future does the forecast go?
90 days by default in this build — a deliberate, honest scope rather than promising indefinite accuracy, since forecasts naturally get less reliable further out.
Can I see the forecast across multiple accounts combined?
Not in this build — it forecasts one account's balance. Blending multiple accounts into one combined forecast is real additional work Pocketsmith has built.
What does it cost to run?
Hosting and a database — typically a few dollars a month, with no external API dependency for the forecast itself.
Related tools
Receipt