Spot best-sellers that refunds make unprofitable before restocking
Have Grok Bot cross-check Amazon sales against Sellerboard so you spot refund-heavy best-sellers before you restock.
What the agent did
Source-reported example · as described by TonyGainzzz (@TonyGainzzz)
An Amazon seller connected Grok Bot to Amazon's API and to Sellerboard, then had it analyze the last 6 months of sales to find ASINs with high sales volume but high refund rates and little or no profit, putting about 35 ASINs into a table.
Five of those were back in stock and he had been about to replenish all of them.
He reports a table of about 35 ASINs; of the 5 he was about to replenish, he bought 1 instead of 5 and says only 1 of the 5 was actually worth replenishing.
Limitations
Short X post. No dollar savings, refund or profit thresholds, run time or prompt were shared, and the table itself is not shown.
The post doesn't say whether the bot or the author judged which of the 5 was worth replenishing; the purchase decision (buying 1 instead of 5) was the author's. He notes the profit check depends on cost-of-goods data from Sellerboard, because Amazon data alone has refunds and fees but no cost of goods or profit.
How the Amazon API connection was set up is not described.
Try it in your business
What you'll need
Grok Bot; Amazon seller API access connected to the bot; Sellerboard (or similar) with cost-of-goods data and a bot user login; several months of sales/refund history. Exact thresholds, table columns, API setup details, and model are not disclosed.
Grok Bot was connected to Amazon's API and given its own user login in Sellerboard; no custom code described.
Starting prompt
The author didn't publish their prompt (checked full note_tweet text of the post and the author's own replies in the X conversation; no prompt quoted), so this is our adaptation.
How ours differs from the original setup: Same Grok Bot execution agent and the same two data sources. Spells out the table columns, asks for the numbers behind each recommendation, marks missing cost data 'unknown', and keeps the bot read-only (no orders); the purchase decision (buying 1 instead of 5) was the author's.
Using my connected Amazon seller data and Sellerboard, look at the last 6 months for every ASIN. Build a table with units sold, revenue, refund count and refund rate, Amazon fees, cost of goods (from Sellerboard) and net profit. Flag ASINs with high sales but a high refund rate and little or no profit. Then check which flagged ASINs are back in stock or due for a reorder and, for each one, say whether it looks worth replenishing and why, showing the numbers. If cost data is missing for an ASIN, write 'unknown' instead of guessing. Don't place, change or cancel any orders, listings or shipments; this is a recommendation for me to review.
Replace the [bracketed] parts with your own files and tools.
Get examples like this every Tuesday and Friday
Practical ways to put AI agents to work in your business. We check the original sources, explain the setup and point out the limits. Free.
Unsubscribe any time. Privacy policy
Related use cases
Build an estimate and job-cost tracker connected to QuickBooks, email and OneDrive
Have Grok Bot build an estimate and job-cost tracker connected to QuickBooks, email and OneDrive, as the author says he did to replace his construction-management software.
Our suggested starting prompt — not the author's original · untested
You'll need: Grok Bot with access to your QuickBooks, email and OneDrive (the post does not say how it was connected); the estimate and job-cost categories your current software uses, ideally with a sample from one job; your own review of the tracker and of anything the bots prepare.
I want an estimate and job-cost tracker to replace [current software or spreadsheet]. Build it as a workbook saved in [OneDrive folder] with: an estimate sheet per job ([line items and cost codes]), a bills and actual-cost sheet per job, and a summary of budget against actual cost per job. Fill the cost data from QuickBooks ([company file]) and from bill emails in [mailbox or label], read-only: do not change anything in QuickBooks, send any email or delete anything. Match each bill to a job and cost code, and list the ones you cannot match, with the reason. First show me the structure and a sample for [one job] before you fill in the rest. For draws: when I ask, prepare the draw package for [job], draw number [n], using only invoices that were not in earlier draws, and list any issues from the last draw request. Save it as a draft and write a draft email to [bank contact]. Do not send anything to the bank; show me the package and the draft first.
Project inventory and recommend when to reorder each product
Have Claude Code pull inventory and sales data, project the next twelve months of sales and stock by product, and recommend when to reorder each one, using written notes on supplier lead times and holidays.
Author's example instructions
You'll need: Claude Code; sales and inventory data by SKU (the author pulled it from Amazon with scripts Claude Code wrote); written notes per brand covering supplier lead times, holiday closures, off-platform stock and past stockouts; an instruction file or command listing the projection steps. Not disclosed: The scripts' code and how the Amazon credentials were connected; the full text of the instruction file and per-brand context files; the number of brands and SKUs; how accurate the projections turned out against actual sales.
1. Check to see if the sales/inventory data we have locally is >5 days old, and if not, pull fresh data from Amazon. 2. Check the context notes for the brand it's working on. 3. Pull inventory and sales data from the local cache. 4. Display the trailing twelve month sales data by SKU and by month. 5. Describe any factors it's going to use to make adjustments to TTM sales data when projecting NTM sales. 6. Review its stated adjustment factors, then project NTM sales by SKU on a monthly basis, and display the amount of stock remaining at the end of each month based on those projections. Use Python for the math here. 7. Recommend when to reorder each SKU, factoring in CNY and other holidays.
Turn manufacturer emails into service calls and draft replies for your approval
Have Claude Code read the mailbox, turn manufacturer emails into service calls and draft the replies, with anything that has consequences waiting for the owner's approval on his phone.
Our suggested starting prompt — not the author's original · untested
You'll need: Claude Code access; a business system that holds service calls (the author built his own ERP) plus an MCP server or similar connection that lets the agent read the mailbox and create records in it; an approval step on your phone or desktop (the author uses Telegram with approve/reject buttons and a decision log); written procedures the agent must read before it writes anything (the author's company has about 1,500); your own review and approval of every service call and reply before it is created or sent
You help me run the office of my equipment service company. Check the mailbox [mailbox] for new emails from the manufacturers we service for [manufacturer names]. Before doing anything, read our written procedures for handling manufacturer requests [procedures file or folder]. For each email that asks for a service visit, prepare a draft service call (do not save it in [our job or service system] until I approve) with the customer site, machine model, fault reported, manufacturer reference and urgency, taken only from the email; mark anything missing or unclear instead of guessing. Then draft a reply to the manufacturer confirming what we understood and asking for any missing details. Put each draft service call and draft reply in my approval queue [where] with a one-line summary. Do not create, change or close any record, send any email, order parts, or delete anything; I will review and approve each one myself.

