Inventory & Ops · Blog post · source May 14, 2026
A Cashflow and Safe-Distribution Dashboard with Claude
Project each brand's cash by month to see how much can be taken out while still funding reorders
It took about half a day on top of the earlier work on the inventory skill.
Per source Self-reported by the source, not independently verified.
Copy-paste prompt
For brand [X]: starting cash $[ ] plus Amazon's pending balance $[ ]. Using my projected monthly unit sales and reorder dates [paste], unit cost, freight per unit and payment terms ([30]% deposit at order, rest on completion), estimate monthly cash in (after referral, fulfillment, storage and returns) and cash out for the next 12 months. Tell me the most I can distribute today while keeping cash above $[buffer] every month. Show the table and your fee assumptions.
The source didn't publish its exact prompt. We wrote this starter so you can try the same workflow. Replace anything in [brackets].
What they did
Alex Willen had Claude build a dashboard showing how much cash he can safely distribute from each of his Amazon brands. It starts from each brand's ledger balance plus the cash Amazon is holding, projects sales and reorder dates with his /inventory skill, and models deposits (30/70 or 50/50), freight, referral and fulfillment fees, storage (FBA vs AWD, Q4), returns and promotion costs. It refreshes one or two brands a day so each brand updates weekly.
How it went
It took about half a day on top of the earlier work on the inventory skill. He says the skill only became reliable with Opus 4.7 at max reasoning (5 of 5 test cases). Along the way he caught Claude reading the wrong cash column, skipping the inventory skill, and double-counting Amazon fees. Once it was right, he paid himself a distribution.
We only report what the source says. Results are self-reported unless noted.
Worth knowing
Add charts and a 'method used' card so you can check the numbers; that's how he found the errors. He keeps a cash buffer per brand because the fee models are simplified.