Running an Amazon Store (Audits, Listings, Reviews, Ads) with Claude Code Skills for Amazon Sellers

What they did

A seller says Claude Code runs his Amazon store through the official seller APIs, using 5 Claude Code skills rather than scripts. The agent re-runs the playbook itself (audit, fix, measure). He lists what that covers: review requests on full autopilot, 19 live listings managed via the API, keyword intel across the catalog (about 700K monthly searches mapped), every change baselined and measured weekly, and ad spend and full campaigns managed through the Amazon Ads API.

How it went

He reports the store sold $130K in about 6 months (6,454 units, 7.7% conversion rate, about 19% net, $250K of stock at list price) and attached a screenshot of the sales figure. He doesn't say how sales compared with before the agent took over.

We only report what the source says. Results are self-reported unless noted.

Worth knowing

Self-reported, with a screenshot. He said he was open-sourcing the setup in a thread, but the repo link isn't in the posts we could open. Review requests through Amazon's Request a Review feature are allowed. Asking only happy buyers for reviews is not.

Copy-paste prompt

Act as my Amazon store operator. Each week:
1. Audit: pull sales, sessions, conversion and ad data per ASIN for the last 7 days vs the 4 weeks before. List the 5 biggest problems with numbers.
2. Fix: for each problem, propose one specific change (listing text, bid, budget, negative keyword). Record today's baseline for the metric it should move.
3. Measure: check last week's changes against their baselines and say which helped, which didn't, and what to roll back.
Wait for my approval before making any change.

The source didn't publish its exact prompt. We wrote this starter so you can try the same workflow. Replace anything in [brackets].