Bring an ASIN
The skill follows accessible review pages, star filters, sort orders and variants. It removes duplicate reviews and records what it could—and could not—collect.
Aspi skills / Amazon research
Amazon reviewAudit
Your next product improvement is probably sitting in a review. Give Codex an ASIN and turn accessible shopper feedback into a report you can actually use.
$9One-time purchase
Reusable Codex skill ZIP
Reusable Codex skill ZIP
Requires your own Codex access with built-in browser tools and Python 3.9+. Amazon limits can prevent a complete review export.
Screenshot of the results summary. View the full report and downloadable data below.
Read the feedback.
Find the next fix.
Go beyond the stars
Find what shoppers value, where they struggle, and what five-star reviews still want improved. Keep delivery issues separate from product problems.
Make a better next version
Get prioritized product-refresh and new-product opportunities, supporting review links, confidence notes, tradeoffs and concrete validation steps.
The Echo Dot Kids test run.
ASIN B09B9CD1YB. Four variants. A real collection with real limits, plus a report showing where the product experience could improve.
958Unique reviews collected
2,081Written reviews displayed
46.0%Collection coverage
Read the example honestly: 1,123 displayed reviews were not collected. The expanded dataset uses rating proxies and preliminary English keyword tags. The six opportunities came from a manually read sample, not semantic analysis of all 958 reviews. The current skill explicitly tracks collection and analysis separately.
Make setup easier
Parents describe trouble connecting child accounts and accessing intended audio. The report proposes a per-child content check and test playback.
Make recovery clearer
Connection and reset complaints suggest testing better diagnostics and recovery. The report does not assume every issue is a hardware defect.
Serve different buyers
Parents and adults buying decorative designs have different needs. The report separates those use cases before proposing new-product concepts.
Read the full reportReport · Markdown958-review data · CSVOpportunities · CSVPreliminary topic counts · CSVCoverage · JSON
Independent research example. Aspi is not affiliated with or endorsed by Amazon. Customer reports and proposed solutions are not independently verified product facts.
03 / Inside the ZIP
Keep the whole workflow.
- Reusable Amazon Review Audit skill
- Start and installation guide
- Deep collection and deduplication instructions
- Product-opportunity report requirements
- Python helper for normalized review CSVs
- Coverage and evidence rules
Start with what buyers said.
One reusable workflow for your products, competitors, and the next idea you want to investigate.
Before you buy.
Will it collect every review?
Not always. Amazon can cap views, hide reviews, require sign-in or change the page. The skill tries meaningful available filters and reports exactly what it collected. It does not promise a complete export.
Is this a finished audit or reusable software?
You are buying a reusable Codex skill folder with supporting instructions and a Python helper. Codex performs the browser work and analysis. This is not a standalone application, API subscription or done-for-you service.
What setup do I need?
Codex with working built-in browser tools, an Amazon browsing session, and Python 3.9 or newer for the normalizer. The helper uses Python’s standard library. If sign-in or CAPTCHA blocks a run, you may need to complete that step yourself.
Can I use it for other products?
Yes. Supply another ASIN or product URL. It is designed for different categories and retains variant information so a shared review family is not mistaken for one exact version.
Does it prove what I should manufacture?
No. It connects proposed improvements to review evidence, then gives you validation steps. Demand, engineering feasibility and commercial impact still need testing.