Get tasks doneBusiness size: Solo

A Malaysian sauce maker hands her Monday numbers to Claude Code, and caught it pricing her costs from the supermarket shelf

Eunice Yin Ern (OTSATE) · Malaysia · Claude Code, Google Sheets

Go to the skill

What the agent did

Eunice Yin Ern runs OTSATE, a small lemongrass sauce brand in Malaysia, and does the cooking and the numbers herself. Every Monday she tells Claude Code last week's sales, ingredient and packaging spend, cash in the bank and the decision she faces. It writes the row in a Google Sheet, works out revenue per day, food cost percentage and weeks of fixed costs covered, and a script it wrote sorts her bank statement PDFs into categories for her to confirm. When it costed a new product with supermarket prices, she added a standing rule that it must never estimate her costs.

Result, as claimed

"What I stopped doing is opening the spreadsheet to type. What I do with that time is read it. Ten minutes, every Monday, and I actually look." On the new product, the AI said contribution of "12 to 19 per cent", but "With my real prices the contribution was 61 per cent."

This is the result as claimed in the public source linked below. It has not been checked and is not a typical outcome.

Proof

Read Eunice Yin Ern's article on e27

Try it yourself

A weekly numbers routine where you say four figures out loud and the AI fills your sheet, with a rule that it never guesses your costs.

Difficulty: Intermediate

Prompt

Standing rule: never estimate my costs. Every cost line is either a price I gave you or marked OPEN, and you make no recommendation that depends on an OPEN line. Ask me instead. This week: sales were [amount], I spent [amount] on ingredients and [amount] on packaging, cash in the bank is [amount], and the decision I am making is [decision]. Add a row to my [sheet name] with these, then work out revenue per day, food cost as a percent of sales and how many weeks of fixed costs [monthly fixed costs] this week's sales would cover. Leave the decision column blank for me.

Prompt: starter by Solopreneur Code

Tools

  • Claude Code Used in the story to update her sheet and sort bank statements
  • Google Sheets Used in the story for the weekly numbers
  • Claude Simpler option if you paste the figures and update the sheet yourself

Steps

  1. Save the standing rule in the memory or project file your AI reads each session, as she did.
  2. Each Monday give it your four numbers and the decision you face.
  3. Confirm any bank transaction category it guesses before it counts.
  4. Read the inputs before the conclusion, and write the decision yourself.

Watch out: AI fills gaps with prices it finds online, which can be double what you pay, so give it your real costs and check every line before you reprice anything.

Based on Eunice Yin Ern (OTSATE)'s steps

Skill: Money and bookkeeping

Add this skill

Money and bookkeepingDifficulty: Intermediate

Weekly numbers check-in

A short weekly routine where you give four figures and the AI fills your sheet, with a rule that it never guesses your costs.

Works with
Claude, ChatGPT, Gemini, Any AI chat The story used Claude Code to write straight into Google Sheets. In a normal AI chat you copy the row into your sheet yourself.
Tools in the story
Claude Code, Google Sheets, Claude
Country
Malaysia

How to use this: Copy the skill, paste it into Claude, ChatGPT or Gemini, and ask it to save it as a skill. If it cannot, it will still follow the steps.

Show skill

If the Copy button does not work, select this text and copy it.

Job: keep the books and catch errors

Other tools for this job

Tools used

  • Claude CodeFree, then Pro $20 a month ($17 a month billed yearly), Max $100 and Ultra $200 a month. Claude Code is included in paid plansChecked 2 November 2026. Price source: Claude Code, opens in a new tab
  • Google SheetsSee official pricing

Could also do this

  • Osome (Singapore) Not tested in a storySingapore firm. Says it pairs an accounting team with its own software for bookkeeping, invoicing, tax and filing deadlines.Page checked 2 November 2026. Vendor page: Osome (Singapore), opens in a new tab
  • Dext Not tested in a storyCaptures and categorises receipts, invoices and expenses and syncs them to your accounting software. The page does not mention checking prices against a list.Page checked 2 November 2026. Vendor page: Dext, opens in a new tab

Sign up for the newsletter

Get the next story from Solopreneur Code in your inbox.

Form not loading? Subscribe on Substack

Back to all stories

More like this

Browse all stories
Get tasks doneBusiness size: SoloIncludes prompt

A newsletter writer's Claude Code agent turned screenshots into 38 scheduled Substack Notes he did not write, with no engagement lift

Dheeraj Sharma @genaiunplugged · Country unknown · Claude Code, SubflowAI Substack connector, macOS scheduler, Substack Notes

Each morning a headless Claude Code run takes the oldest screenshot, classifies it, writes a 45 to 115 word Note in his voice, checks it against copy gates, uploads the image, schedules it for the next free day, reads the schedule back and archives the screenshot. Screenshots with emails, phone numbers or orders are set aside.

Result, as claimed

29 Notes scheduled by 23 Sep and 38 by 12 Oct with none written by him. 12 measured Notes had a median of 3.5 likes against 5 for his other Notes, which he reads as no lift.

Get tasks doneBusiness size: CompanyIncludes prompt

TripleTen says Make scenarios cut its Meta ad setup time by 5 to 7 times and label 1,000+ creatives a year with AI

Maksim Epifanov · United States · Make, Make AI Agent, Gemini, Google Sheets, Meta Ads, TikTok, Google Ads, Apify, Slack

The agent collected competitor ad examples and AI labelled creative assets. A separate fixed Make scenario created ad structures and uploaded approved assets.

Result, as claimed

Make reports a 5 to 7 times cut in time spent on Meta ad setup, and creative labelling of 1,000+ assets a year going from weeks to seconds. The 1,000+ figure refers to 2025 volume.

Get tasks doneBusiness size: Small teamIncludes prompt

Layer says Claude and Supermetrics cut a client report from 10 hours to 20 minutes

Morten Kleven · Norway · Claude Cowork, Supermetrics connector, Google Sheets

Claude queried normalised live marketing data via Supermetrics and assembled report analysis.

Result, as claimed

Anthropic's case says client reports that took 10 hours of manual spreadsheet work now take 20 minutes, and 17 of those minutes go into dropping the finished data into the client's Google Sheet.