Amazon MCP server
{
"mcpServers": {
"trendsapi-amazon": {
"type": "http",
"url": "https://amazon.api.trendsapi.ai/mcp"
}
}
}{
"mcpServers": {
"trendsapi-amazon": {
"type": "http",
"url": "https://amazon.api.trendsapi.ai/mcp"
}
}
}{
"mcpServers": {
"trendsapi-amazon": {
"type": "http",
"url": "https://amazon.api.trendsapi.ai/mcp"
}
}
}{
"servers": {
"trendsapi-amazon": {
"type": "http",
"url": "https://amazon.api.trendsapi.ai/mcp"
}
}
}{
"mcpServers": {
"trendsapi-amazon": {
"type": "http",
"url": "https://amazon.api.trendsapi.ai/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 1.0.1
What it does
This MCP server provides access to Amazon product search interest and bestseller feed data via the Trends API. It exposes methods for retrieving historical search volume time series, growth calculations over configurable windows, and live bestseller rankings. It is intended for e-commerce researchers and developers who need demand signal data without building scrapers.
- Retrieve a weekly time series of Amazon search interest for a product keyword to track demand over time.
- Calculate percentage growth in search interest for a keyword over multiple periods (e.g. 3M, 12M, YTD) to identify trending products.
- Fetch a ranked list of current Amazon Best Sellers or Top Rated products, optionally filtered by category.
Tools this server exposes
| Tool | Does |
|---|---|
| get_time_series | Returns a list of dated search-interest index points (0–100) for a given keyword on Amazon or another supported source. |
| get_growth | Returns growth percentage and direction for a keyword over one or more specified time windows. |
| get_live | Returns a paginated ranked list from a live Amazon bestseller or top-trends feed, with optional category filtering. |
| get_top_trends | Returns a top-trends feed by specifying an exact feed type such as 'Amazon Best Sellers Top Rated'. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.