Real Agent Cases

Run a restaurant's back office through an OpenClaw agent: supplier price-change alerts, invoice uploads and a fairer tip split

Posted
by Christopher Thépot (LinkedIn)
Last checked
Method:
How we verify

Evidence: Described by posterThe person describes what happened in words, without screenshots or video.

Outcome

Christopher Thépot runs restaurant operations in Madrid and spent a month using OpenClaw. The agent answers point-of-sale questions over WhatsApp and combines supplier invoices, sales data, recipes and inventory counts to track waste and what's missing. It flags supplier price spikes; two pricing mistakes caught so far directly affected the P&L. It also pulls supplier invoices from their inbox into the accounting software's scanner, saving one to two hours a week, though they're still fine-tuning that. They used it to design and test a new tip policy in under an hour: 80% stays with the shift and 20% goes to a pool split by hours worked.

Agent

OpenClaw

Prompt Recreated prompt

The source didn't share its exact prompt, so we wrote this from what it describes.

Before you start, ask me 1–3 short questions about anything important that's missing below. Keep them quick. Then do the task, checking in with me at the key points.

Goal: Catch supplier price changes and invoice mistakes for my restaurant before they hit my margins.

What I can give you (ask for whatever's missing):
- Supplier invoices (PDF or email): [UPLOAD]
- My agreed prices or last month's prices per item: [PRICE LIST]
- How big a change should alert me: [THRESHOLD]

Check any change to staff pay or tips against local labour rules.

Rules:
- Read every line: item, unit, quantity, unit price.
- Compare with the last price paid for the same item and unit, and flag changes over my threshold.
- Flag maths errors, duplicate invoices and items I don't remember ordering.
- Draft a short, polite query to the supplier for each problem; don't send it.
- If a site asks me to sign in, pause and let me do it myself (including any CAPTCHA or two-step code); never guess or store passwords.
- Check facts that can change (prices, opening hours, dates, availability, rules) against the official source, and tell me anything you couldn't verify.

Output: A weekly table of price changes and errors, with the cost impact and a draft supplier message for each.

You'll know it worked when: No price change or invoice mistake goes unnoticed for more than a week.

Try this in:

This prompt has blanks to fill in, like [UPLOAD]. We'll copy it instead of filling it in, so you can edit it before you send.

The source used OpenClaw, which can take actions in a browser or app. These are general chat tools, so results may differ, and a normal chat may only give you a plan.

Opens in a new tab. We don't see what you send.

These links are only a shortcut. We're not affiliated with OpenAI, Google or xAI, and your results may not match the source.

Optional stronger version

Our rewrite with extra checks. The poster didn't test it.

Before you start, ask me 1–3 short questions about anything important that's missing below. Keep them quick. Then do the task, checking in with me at the key points.

Goal: Catch supplier price changes and invoice mistakes for my restaurant before they hit my margins.

What I can give you (ask for whatever's missing):
- Supplier invoices (PDF or email): [UPLOAD]
- My agreed prices or last month's prices per item: [PRICE LIST]
- How big a change should alert me: [THRESHOLD]

Check any change to staff pay or tips against local labour rules.

Rules:
- Read every line: item, unit, quantity, unit price.
- Compare with the last price paid for the same item and unit, and flag changes over my threshold.
- Flag maths errors, duplicate invoices and items I don't remember ordering.
- Draft a short, polite query to the supplier for each problem; don't send it.
- If a site asks me to sign in, pause and let me do it myself (including any CAPTCHA or two-step code); never guess or store passwords.
- Check facts that can change (prices, opening hours, dates, availability, rules) against the official source, and tell me anything you couldn't verify.

Output: A weekly table of price changes and errors, with the cost impact and a draft supplier message for each.

You'll know it worked when: No price change or invoice mistake goes unnoticed for more than a week.

Before you finish: re-read my goal and constraints, check your answer against each one, and list anything you assumed or couldn't confirm.

Source

https://www.linkedin.com/posts/christopher-th%C3%A9pot-03217391_openclaw-activity-7452697032458334209-S_6n

  • Michel, in San Diego, works in a tight-margin business with basic back-office software, and sending late-invoice reminders was repetitive manual work. Claude Cowork walked them through setting up an agent that logs in every Monday night, checks overdue accounts receivable and matches customer names against a Google Sheet of contacts. Using Michel's email template, it adds the contact name, invoice number, amount and due date, puts the emails in the Drafts folder and sends Michel a report of which customers were included, what each owes and the totals. Michel reviews the drafts and sends them, saving about an hour every Monday night.

    Claude Cowork · Ops · Source date:

  • mattmaroon runs a seasonal business and noticed their state unemployment insurance rate had jumped from 0.75% to 12.75%. Instead of hours on the state website and on hold, they asked ChatGPT what had happened. It worked out that they were being charged a penalty rate because their new payroll provider doesn't file wage reports for the one quarter with no wages, and without a report the business counted as delinquent. ChatGPT filed the correct wage report and went to the support site to request a retroactive rewrite, which was granted, saving many thousands of dollars without hitting the 5-hour usage limit.

    ChatGPT · Ops · Source date:

  • A client called Alper Ortac on a family holiday at Europa Park needing an invoice urgently for their end-of-year accounts. With no laptop, Ortac opened Claude on their phone in a roller coaster queue, attached the last invoice and asked it to recreate it. Claude wrote and ran a Python script that produced a replica of the PDF, then changed the invoice number, billing period and costs as asked. Ortac drafted the email, attached the new invoice and sent it in roughly 20 minutes, right before boarding.

    Claude · Ops · Source date: