---
title: "Receipt processing in Cowork"
canonical: "https://agenticfeats.com/c/receipt-processing-in-cowork-claude-cowork/"
agents: ["Claude Cowork"]
evidence_grade: "C"
outcome: "partial"
job_type: "Finance and Admin"
industry: "Universities"
roles: ["Operations"]
source_url: "https://sites.duke.edu/ddmc/2026/02/24/what-claude-cowork-actually-does-and-what-it-doesnt"
source_date: "2026-02-25"
published: "2026-10-13"
updated: "2026-10-13"
---

> From Agentic Feats: sourced examples of AI agents at work. Map of the site for agents: https://agenticfeats.com/llms.txt

# Receipt processing in Cowork

Claude Cowork processed a PDF purchasing card receipt, extracting data and producing a cover sheet for the author's TPS report, running its own Python script.

Result: Cowork ran the script, fixed an error itself, and returned finished output files with no terminal or manual dependency setup from the author.

## How it was set up

Claude Cowork desktop research preview, pointed at a local folder of PDFs. It runs a Linux VM locally with reasoning done in Anthropic's cloud.

## Worth knowing

The session crashed after a brief connectivity loss and needed the app's stored data cleared. Sessions are tied to one machine and not shared.

## Prompt

```text
Read the receipts in [folder], extract the key data, organize it into a spreadsheet, and produce a cover sheet for my [report name].
```

## Evidence note

Author's firsthand blog account; no number or output file shown

## Sources

- Original: https://sites.duke.edu/ddmc/2026/02/24/what-claude-cowork-actually-does-and-what-it-doesnt

## The receipt

- Agent: Claude Cowork
- Evidence: Claim (grade C, claim only)
- Outcome: Partial
- Job type: Finance and Admin
- Industry: Universities
- Roles: Operations
- Source: Blog
- Published: 2026-02-25
- Added: 2026-10-13