MCP server · 3 tools · 4 endpoints

Amazon MCP server

Product pages, keyword search, customer reviews and seller offers from Amazon, delivered to Claude or Cursor without a browser in the loop.

Get a free key Connect in a minute

100 credits on signup · No card · Sign in with OAuth or paste a key

scrapersocial-amazon mcp.scrapersocial.com/amazonconnected

you

Pull the 50 most recent reviews for <amazon product url> and group the complaints that keep coming up

agent

calls get_reviews 3 credits / review

{ "url": "<Amazon review url>" }

server

returns

reviews [{ rating }] · one per review

agent

Reads the reviews and groups the complaints that keep coming up.

Shape only. Keys follow the tool's summary; a real call returns the full record. 1 call · 3 credits

Connect

One URL. Your client already knows what to do with it.

The first call opens a sign-in. New accounts get 100 credits, no card. Every client below shares one account and one balance.

Get a free key

server url https://mcp.scrapersocial.com/amazon

claude.aiOAuth

Settings → Connectors → Add custom connector, then paste the server URL.

mcp.scrapersocial.com/amazon

Web, desktop and mobile. Nothing else to paste.

Claude Codeone command

Run this in a terminal.

$ claude mcp add --transport http scrapersocial-amazon https://mcp.scrapersocial.com/amazon

Approve the sign-in prompt on first use.

Cursorone click

Installs the server and opens the MCP panel.

Add to Cursor

or paste the JSON into .cursor/mcp.json

{
  "mcpServers": {
    "scrapersocial-amazon": {
      "url": "https://mcp.scrapersocial.com/amazon",
      "headers": {
        "Authorization": "Bearer sk_live_..."
      }
    }
  }
}

Other clients · VS Code, Windsurf, Gemini CLI, anything that speaks MCP

VS Codeone click or JSON

Add to VS Codeor run MCP: Add Server, choose HTTP and use:

{
  "name": "scrapersocial-amazon",
  "type": "http",
  "url": "https://mcp.scrapersocial.com/amazon"
}

Windsurf~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "scrapersocial-amazon": {
      "serverUrl": "https://mcp.scrapersocial.com/amazon",
      "headers": {
        "Authorization": "Bearer sk_live_..."
      }
    }
  }
}

Gemini CLI~/.gemini/settings.json

{
  "mcpServers": {
    "scrapersocial-amazon": {
      "httpUrl": "https://mcp.scrapersocial.com/amazon",
      "headers": {
        "Authorization": "Bearer sk_live_..."
      }
    }
  }
}

Any clientStreamable HTTP · OAuth 2.1 or Bearer

$ npx add-mcp https://mcp.scrapersocial.com/amazon

One account and one balance across every client and the every-platform hub at mcp.scrapersocial.com.

What it does

What your agent can do with Amazon

Say you look after pricing for forty SKUs and want to know, every morning, what each one sells for on Amazon and who else is offering it. That’s a job for an agent, and the ScraperSocial MCP server for Amazon gives it the four reads it needs: the product page, keyword search, customer reviews and the list of seller offers on a listing.

Amazon isn’t a social network, and this page doesn’t pretend otherwise. The server treats it as public web data, meaning whatever a logged-out visitor can see on a listing is what your agent gets back, structured, over the same connection it already uses for the social platforms.

The spoke exposes one capability tool, get_reviews, which takes a product URL and pages through reviews with their ratings. Everything else is reached through call_endpoint by path: /v1/amazon/product for the listing itself (title, price, rating, specifications), /v1/amazon/product-search for keyword results, and /v1/amazon/sellers for the offers on a product. list_endpoints prints the four with inputs and costs, and calling it is free.

Point your client at https://mcp.scrapersocial.com/amazon, sign in or paste an API key, and ask in plain language. The agent picks the path.

Four jobs it finishes on its own

Daily price tracking

Hand the agent a list of product URLs and have it call `/v1/amazon/product` for each one, then write the price and rating into a sheet. Run it every morning and you get a price history without maintaining a scraper. Add `fresh: true` on the days the exact current price matters.

Review mining on a product you compete with

Pull a few hundred reviews with `get_reviews` and ask the agent to group the complaints. Recurring one-star themes are the fastest signal of what to fix, or what to say on your own listing. Each review costs 3 credits, so set a limit before you start.

Who else sells this

`/v1/amazon/sellers` returns the offers on a listing: the sellers, their prices and conditions. Use it to spot an unauthorised reseller undercutting you, or to see how crowded a listing is before you enter it.

Keyword landscape

Search `/v1/amazon/product-search` for the term a customer would type and get the products that come back, ranked as a shopper sees them. Compare the top results on price and rating to decide whether a category is worth entering.

Limits · read before you connect

What this does not do

Written by a person, not generated. If something here is wrong, tell us.

Public listing data only. There is no login, no cart, no ordering and nothing from Seller Central or the Selling Partner API. There’s no endpoint for a storefront’s full catalogue; the four reads above are the whole surface. Every Amazon endpoint costs 3 credits per item returned, and a call that returns nothing is charged 1. The list calls (search, reviews, sellers) hand back at most 50 items each time, plus a cursor that fetches the following page. Responses may come from cache; fresh: true bypasses it at the same price. Prices are what the page said when it was read, not a live quote.

Tools

3 tools, same names as the every-platform hub

One does one Amazon thing each. The last two reach all 4 endpoints by path, so nothing in the catalog is out of an agent's reach.

Full credit table →

Read reviews

get_reviews

Public reviews and ratings for a product, business, app or company page.

3 credits / review

Reaches the other 3 endpoints

List every Amazon endpoint

list_endpoints

Browse every Amazon endpoint, its input and its credit cost. Free, and calls nothing upstream.

free

Call any Amazon endpoint by path

call_endpoint

Run any Amazon endpoint by its path, so nothing Amazon offers is out of reach of an agent.

endpoint's own cost

Pricing

100 free credits is 33 reviews.

Then $5 a month for 1,000 credits, or $4 per 1,000 as a top-up. No card to start. Every call costs 3 credits.

33 reviewsget_reviews · 3 each

Get a free key 100 credits on signup · $5 for 1,000 · full credit table

Prompts

Things to ask once it is connected

Copy one, swap the placeholder, paste it into your client.

Get the current price and rating for <amazon product url>

Pull the 50 most recent reviews for <amazon product url> and group the complaints that keep coming up

Search Amazon for <keyword> and list the top 10 results with price and star rating

Who is selling <amazon product url> and at what price? Flag anyone below <price>

Compare these three listings on price and review count: <first listing url>, <second listing url>, <third listing url>

Questions

Before you connect

What does one Amazon call cost? #

Three credits per item. A product page is one item, so 3 credits. Twenty reviews are 60 credits. A search returning 10 products is 30. Signing up gives you 100 credits free, and the next 1,000 cost $5.

Can the agent buy something or check my orders? #

No. The server reads public listing pages and nothing else. It never signs in, so carts, orders, wishlists and seller dashboards are out of scope.

Is this an official Amazon API, or affiliated with Amazon? #

Neither. ScraperSocial is independent, with no relationship to Amazon. Account-tied data belongs to Amazon's own Selling Partner API.

Which inputs does each endpoint take? #

`/v1/amazon/product`, `/v1/amazon/reviews` and `/v1/amazon/sellers` take a product URL. `/v1/amazon/product-search` takes a text query. `list_endpoints` reports this per path.

How many reviews can I get in one go? #

Fifty per call. The response includes a cursor; pass it back to fetch the next fifty. Credits are charged per review returned.

Go further

If your job is knowing what things cost on Amazon and what buyers say about them, this spoke does it without a scraper to babysit. Add https://mcp.scrapersocial.com/amazon to whichever MCP client you use, start on the 100 free credits, and let the agent choose the endpoint.

Get a free key 100 credits, no card

ScraperSocial is independent and not affiliated with, endorsed by or connected to Amazon. Publicly available data only.