Claude to package quarterly accounting evidence
Remco Livain wanted a quarter's accounting records to arrive at his accountant without another hunt for missing receipts. His setup had been developed over months using Claude Code and Cowork. It accepted scanned letters, PDF documents, email receipts and photographs, classified them, and maintained a receipt register against bank and card transactions.
livain.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Match this quarter's bank and card statement lines to the receipt register. Flag missing evidence, then create one PDF per account with the cover checklist, official statement and referenced receipts.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Remco Livain
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Livain reports a first quarter with zero missing receipts and roughly 45 minutes to produce the requested PDF packages.
- Human involvement
- Livain refined the workflow over months; the accountant supplied classification and handoff rules.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Collect statements and receipts into the established capture workflow.
- Match every statement line against the receipt register and flag gaps.
- Assemble one indexed PDF per account and quarter using the accountant's format.
Remco Livain wanted a quarter's accounting records to arrive at his accountant without another hunt for missing receipts. His setup had been developed over months using Claude Code and Cowork. It accepted scanned letters, PDF documents, email receipts and photographs, classified them, and maintained a receipt register against bank and card transactions.
The final handoff followed the accountant's requested format: one PDF for each account and period. Each package contained a cover sheet and checklist, the official statement, and receipts in chronological order. Livain used prompts and scripts to sort, rename and stamp references so a transaction and its evidence could be connected. Missing evidence was flagged explicitly rather than disguised as a complete record.
He says the packaging took about 45 minutes and that this was his first quarter without a missing receipt. That result followed refinement of a recurring process, rather than an instant solution from a single prompt.
The accountant supplied the rules for business versus mixed-use cards and for the relevant payment and VAT treatment. Livain wrote those decisions into the workflow so the same choices did not have to be reconstructed every quarter. The concrete output was an organized evidence packet that matched the recipient's expectations.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: www.livain.com · Remco Livain
Frequently asked questions
What was inside each handoff PDF?
A cover sheet and checklist, the official account statement, and chronologically ordered receipts with matching references.
How were missing receipts treated?
The workflow flagged absent evidence explicitly instead of presenting an incomplete package as complete.
Which decisions came from the accountant?
Rules for business versus mixed-use cards, payment treatment and VAT handling were captured for reuse.
More details
About this account
Claude classified evidence and helped scripts reconcile, sort and package the quarterly records.
Firsthand source opened and read. Suggested prompt written by us from the reported workflow; no original task prompt is quoted.
Credit: Remco Livain
Source published
Added to Agents In Use
Source checked
Tools and topics
Claude Code · Claude Cowork · PDF statements · Receipt register
Receipt reconciliation · Accountant handoff · Quarterly records
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.