Manus to turn family receipts into tax workbooks
Keitaro_aigc used Manus to prepare Japanese tax paperwork from sixty three medical receipts and ten hometown tax donation certificates. They photographed the documents, packaged the images in ZIP files and asked for spreadsheets, dated PDFs and consistently renamed source images.
Original prompt
Shared by the source author. See source notes for context.
次は確定申告用の様式に合わせて
Reach people exploring AI agents
See placements →What happened
The reported fifteen minute preparation produced medical and donation workbooks, PDFs and renamed images; filing and paid refunds were not verified.
Reported by keitaro_aigcHow the agent did it
- Upload medical receipts and donation certificates as image ZIP files.
- Correct patient names and tax years before transferring data into the official workbook.
- Generate dated PDFs, renamed images and fiscal year folders with a README.
Where a person stepped in
The author corrected misread patient names, years and image orientation before accepting the prepared documents.
Keitaro_aigc used Manus to prepare Japanese tax paperwork from sixty three medical receipts and ten hometown tax donation certificates. They photographed the documents, packaged the images in ZIP files and asked for spreadsheets, dated PDFs and consistently renamed source images.
The choice of receipt images addressed a limitation in their own records: private dental treatment was not reflected in their usual portal, and collecting separate family permissions there was cumbersome. The image workflow combined the family's documents in one processing task.
Manus read dates, hospitals, amounts and patient names, but some names and years were wrong. The author corrected it with a closed set of four family names and an instruction that all receipts belonged to 2025. They also requested image rotation correction. Manus then searched the National Tax Agency website, downloaded the medical expense Excel template and generated Python to populate it.
The deliverables included a populated official workbook, a date ordered PDF and renamed images. The donation certificates received corresponding spreadsheet and PDF outputs. Manus also organized the files by fiscal year with a README for later GitHub storage. The author reported about fifteen minutes including corrections. This was document preparation; the article does not verify a submitted return, refund payment or error free extraction.
Prompt
Original prompt
Source: note.com · keitaro_aigc
Frequently asked questions
Why use receipt images rather than the portal?
The author's private dental receipts were absent from their usual portal, and separate family permissions added work.
What errors required correction?
Patient names and years were misread; the author supplied four allowed names and the 2025 year constraint.
Was a refund paid?
The article reports generated preparation files, without evidence of submission or a paid refund.
More details
About this account
Manus extracted receipt data, downloaded the official Excel format, generated Python and packaged organized spreadsheet, PDF and image outputs.
Firsthand source opened and read. Prompt is a verbatim excerpt of the published original instruction.
Credit: keitaro_aigc
Source published 12 Mar 2026
Source checked 14 Oct 2026
Tools and topics
Manus · Python · Excel · PDF · GitHub
Medical receipts · Donation certificates · Tax workbook · Image renaming
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.