About this tool
Fill-in-the-blank AI prompts for reconciliations, variance analysis, client emails and plain-English report notes.
The Accountant Prompt Pack is a library of 9 fill-in-the-blank AI prompts for the recurring work of practice and industry accounting: reconciliation difference triage, journal narratives, variance analysis, plain-English report explanations, fee and scope emails, records chasing, close checklists and control reviews. Every prompt carries the same standing rule — the model works only from the figures and rule text you paste, shows its arithmetic, and flags judgement calls as judgement instead of inventing thresholds from memory. You fill the blanks in your browser and copy the finished prompt into any assistant.
Open Accountant Prompt Pack on AltFTool — it loads instantly in your browser.
Add your input to the workspace.
Adjust the options until the result looks right.
Copy or download the output and put it to work.
Each prompt instructs the model to compute from supplied numbers, show the arithmetic, and answer 'requires investigation' where attribution is impossible.
The tax-rule explainer works strictly from official wording you provide, because a model's remembered thresholds may be outdated or from the wrong jurisdiction.
Prompt assembly happens in the browser; no account, no API key and nothing you type leaves the page.
Only with anonymised data and within your firm's and regulator's policies. These prompts are designed to work from figures, rule text and situations rather than names — but you remain responsible for what you paste, so strip client identifiers and check the assistant's data-handling terms before using real balances.
A difference divisible by 9 suggests a transposition error (writing 4,806 as 4,086 creates a difference of 720, which is 9 x 80), and a difference divisible by 2 suggests an item posted to the wrong side — so you search for a single item of exactly half the difference. The reconciliation prompt applies both tests to your exact figure before generating the wider search checklist.
No. It is drafting and analysis support, and every prompt in this pack says so explicitly — journal entries note where judgement is involved, the rule explainer keeps formal advice inside your engagement, and material treatments should be signed off by a qualified reviewer under your normal review process.
Answer their actual worry first, then connect the numbers as a cause-and-effect story with each claim tied to a figure. The plain-English prompt enforces this order, anchors every percentage to an absolute amount, allows at most one analogy, and keeps the whole explanation under 250 words so it survives being read aloud in a meeting.