Perkusai (US)
AI shopping search: real, in-stock US and Amazon products with prices, ratings and buy links.
我該用這個嗎
品質與安全性
根據工具定義與協定合規性的自動化分析。
上下文成本
這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高，可用於其他工作的注意力就越少。
安裝
一鍵安裝
將以下內容加入你的 `claude_desktop_config.json` 檔案：
{
"mcpServers": {
"search": {
"url": "https://mcp.perkusai.com/"
}
}
}遠端端點
https://mcp.perkusai.com/streamable-http它能做什麼
工具清單
工具（1）
🟢search_products(query, k, lang)
Search Perkusai for real, in-stock products in amazon.com (prices in US dollars). Describe what the shopper wants in natural language - e.g. "a full-size basketball", "espresso machine under $200", "wireless earbuds for running". Returns matching products with title, price, rating, stock, market and an affiliate `buy_url` (present it as the purchase link - the price is unchanged for the buyer), plus a `guide`, `coverage` and a paste-ready `markdown` block. Args: query: The shopper's request, in natural language. k: How many products to return (1-12). lang: Response language hint. This catalogue is served in English.
輸入結構描述
{
"type": "object",
"properties": {
"query": {
"description": "The shopper's request, in natural language.",
"title": "Query",
"type": "string"
},
"k": {
"default": 5,
"description": "How many products to return (1-12).",
"maximum": 12,
"minimum": 1,
"title": "K",
"type": "integer"
},
"lang": {
"default": "en",
"description": "Response language hint. This catalogue is served in English.",
"title": "Lang",
"type": "string"
}
},
"required": [
"query"
],
"title": "search_productsArguments"
}輸出結構描述
{
"type": "object",
"additionalProperties": true,
"title": "search_productsDictOutput"
}社群
證據