Review miner · Recipe 2.7
What buyers complain about, in their own words. Lazada reviews grouped by star rating: the one and two star band is a defect list, the five star band is positioning language you did not have to write.
01/ 07
Give it to Claude, ChatGPT or Cursor along with your key. It reads the docs, follows the recipe, and asks you before spending anything.
Read https://data.magpieiq.com/llms.txt and the docs at https://data.magpieiq.com/docs. Then fetch https://data.magpieiq.com/skill/SKILL.md and follow recipe 2.7. My Magpie API key is: <paste your key> Rules: submit only the URLs or items I give you; poll with backoff rather than tight-looping; write the raw payload to disk before analysing it, because re-running a job costs credits again; report anything that failed rather than quietly returning a shorter table; ask me before spending any credits. What I want: read up to 50 reviews on this Lazada product and tell me what buyers actually complain about. Group by star rating, summarise the recurring themes in the one and two star reviews, and quote a few real lines. Tell me how many you read out of the total.
02/ 07
Reviews on the product, read five per call and paged with pageNo
Recurring complaints in the one and two star band
Recurring praise in the five star band, in the buyer's own phrasing
How many reviews were read out of reviewTotal, so the summary can be weighed
03/ 07
/v1/lazada/product-reviews with {"requests":[{"url":"https://my.lazada.co.id/pdp/review/getReviewList?itemId=5133872551"}]}, then poll GET /v1/lazada/retrieve/{task_id}.04/ 07
Real product-reviews response · one Lazada listing · 2 of 5 reviews on page 1 · reviewer identity removed
{
"reviewTotal": 613,
"reviews": [
{
"rating": 5,
"reviewContent": "Bonus pick adalah tambahan yang bagus., Terkesan dengan kerajinan secara keseluruhan., Suka dengan berbagai pilihan warna., Kualitas Yamaha custom terbaik., Pengiriman cepat, menerimanya dalam waktu singkat., Nilai bagus untuk harganya., Desain yang indah, suka warna hitam., ",
"reviewedAt": "2026-07-23T17:00:00.000Z",
"skuFields": [
{
"name": "Warna",
"optionName": "Putih"
}
],
"isPurchased": true
},
{
"rating": 5,
"reviewContent": "okeeeh..semoga awet ",
"reviewedAt": "2026-05-20T17:00:00.000Z",
"skuFields": [
{
"name": "Warna",
"optionName": "Biru muda"
}
],
"isPurchased": true
}
]
}05/ 07
Use the item ID, not the product link. We tried the link first and got a 422, so pull the number from …-i5133872551.html and drop it into the getReviewList URL. On product-reviews-v2, send itemId as a number, not text.
Each call brings back five reviews, and every extra page is its own request. Decide how many pages you want before you start, rather than looping to the end.
These are Lazada reviews only, so describe them that way rather than as sentiment across every marketplace.
Quote a few real lines as buyers wrote them, and mention how many reviews you read out of reviewTotal.
06/ 07
Use this when
Review miner
You have a Lazada listing, yours or a rival's, and want its reviews grouped into themes.
Use instead →
Category share tracker
This reads one product's reviews on one marketplace. It is not a brand health measure and not a substitute for category data — for share and trend, use the Category share tracker.
New accounts start with free trial credits. You generate your own key from the dashboard.