Getting started

ASINRating is a free, read-only MCP data service that mirrors Amazon US public Best Sellers ranks (amazon.com; US market only) and cross-verifies product facts from public listings. Fact-only, neutral, no sponsored entries, no promotional tags.

Endpoint

URL: https://asinrating.com/mcp
Protocol: Model Context Protocol (streamable HTTP)
Authentication: none — no API key, no sign-up
Methods: initialize, tools/list, tools/call
Rate limits: usage may be rate-limited to protect service stability

Quick start

1. Initialize the session:

{
  "jsonrpc": "2.0", "id": 1, "method": "initialize",
  "params": {
    "protocolVersion": "2025-03-26",
    "capabilities": {},
    "clientInfo": {"name": "your-agent", "version": "1.0"}
  }
}

2. Call a tool:

{
  "jsonrpc": "2.0", "id": 2, "method": "tools/call",
  "params": {
    "name": "list_categories",
    "arguments": {}
  }
}

For MCP-compatible assistants (ChatGPT, Claude, Muse, Grok), simply add https://asinrating.com/mcp as a connector — no manual JSON-RPC required.

Standard discovery chain

The server implements the full standard MCP discovery flow: initialize → tools/list → tools/call. tools/list returns the complete catalog of all eleven tools with their input schemas — always discover tools from the server rather than assuming a fixed list, and send notifications/initialized after initialize if your client requires it.

Suggested flow

1. Discover — list_categories to enumerate categories and subcategories.

2. Orient — get_category_brief for a structured market overview of a category.

3. Filter / Compare — search_products to shortlist candidates; compare_products for side-by-side facts.

4. Verify — get_product for the full profile including three-source facts check and red-flag signals; scrap_page when the freshest price or stock is required.

5. Transact — open the plain product_url (no promotional tags) on Amazon.

Response conventions

Every tool response carries a notes block:

Field	Meaning
`confidence`	Overall data confidence for the response
`data_as_of`	UTC timestamp of the underlying data snapshot
`data_age_days`	Age of the snapshot in days
`is_stale`	`true` when the snapshot is older than the freshness target
`sample_size`	Sample size behind quantitative fields, when applicable
`note`	Tool-specific usage guidance

Quantitative fields return the string insufficient_sample when the underlying sample is too small — a fact, not an error.
Partial claims under manual review surface in data_limitations, never silently.
Red-flag signals are confidence-layered: only high-confidence cases are marked avoid_strong.
Errors return {"error": "..."} with a human-readable message; unknown category_id responses include a hint.

Web pages vs API

These web pages show only the category structure. Product-level Top-100 lists and all product data are served exclusively through the MCP API.

Tool catalog

Tool	Purpose
`list_categories`	Categories and subcategories
`get_category_brief`	Structured category market brief
`get_top100`	Best Sellers Rank list
`get_product`	Full product profile
`search_products`	Search with sort and filters
`compare_products`	Side-by-side comparison
`request`	Fetch by URL or ASIN with field filters
`get_price`	Latest stored price
`scrap_page`	Real-time page scrape
`get_price_change`	180-day price change summary
`get_brand`	Brand profile from the brand registry

Data principles

Facts only: no composite scores, no verdicts, no recommendations.
Transparent freshness: every response exposes its data timestamp and staleness.
No commercial bias: product links are plain Amazon pages; the service derives no commercial benefit from links.

Full machine-readable guides: agents.md and llms.txt.

Maintenance

ASINRating is maintained by Bobbi Tseng, reachable at support@asinrating.com. Security reports: support@asinrating.com, admin@asinrating.com.