Linkfox Amazon Search

Amazon Product Search

This skill guides you on how to perform Amazon storefront search simulations, helping Amazon sellers retrieve real-time search result data including product rankings, prices, ratings, and more.

Core Concepts

This tool simulates a real user searching on Amazon's storefront. It returns live search result page (SERP) data: product listings with their positions, prices, ratings, review counts, brands, delivery info, sponsored flags, and more. This is real-time data directly from the Amazon frontend, not historical analytics.

Key distinction from ABA data: ABA data is aggregated historical search term analytics. This tool returns the actual product listings a user would see when searching a keyword on Amazon right now.

Keyword language: Keywords should be in the language of the target marketplace. For example, use English keywords for amazon.com, German keywords for amazon.de, Japanese keywords for amazon.co.jp, etc.

Parameters

Supported Marketplaces

Default marketplace is amazon.com. Use amazon.com when the user doesn't specify a marketplace.

Sort Options

调用方式

• API 端点：POST /amazon/search（完整参数/响应/错误码见 references/api.md）
• Python 脚本：python scripts/amazon_search.py '<JSON 参数>' [--inline]
• 成本约束：本工具会消耗算力；同一会话同一参数组合默认只调用一次，脚本带 24h 本地缓存。失败/空结果不得自动换关键词、翻页或改邮编连续试探；需要继续检索时先向用户说明会产生额外消耗。

输出策略（脚本默认行为）：

• 始终将完整响应写入 <cwd>/linkfox/<YYYY-MM-DD>/<session>/data/linkfox-amazon-search-<timestamp>.json（<cwd> 为脚本执行时的工作目录，在 Claude Code 里即当前项目目录；<session> 取自环境变量 SESSION_ID，按用户任务自动聚合；禁止写入 /tmp，当前目录不可写则报错）
• 响应体 ≤ 8 KB：落盘后把完整 JSON 打印到 stdout
• 响应体 > 8 KB：落盘后 stdout 只输出摘要（顶层字段、常见计数如 total/costToken、最大列表字段的长度 + 前 3 条样本）
• 加 --inline 强制全量打印到 stdout（同样落盘）

读数据建议：先看摘要判断是否足够；需要具体字段时优先用 jq或ConvertFrom-Json 从保存的 json 文件按需抽取，避免整份 JSON 进入上下文。

解决认证和算力问题

发生以下异常情况时，采用 references/onboarding.md 引导解决问题：

异常情况

• 未配置API Key：环境变量未配置 LINKFOX_AGENT_API_KEY，也未配置 LINKFOXAGENT_API_KEY。
• 响应401或402状态码
• 响应提示算力或余额不足：消息含"算力余额不足/计费不足/余额不足/quota exceeded/insufficient balance/套餐到期/需充值/请充值"，或类似含义的内容。

How to Build Queries

Construct the request parameters based on the user's intent:

1. Determine the marketplace: Map the user's target country to the correct amazonDomain value
2. Set the keyword: Translate the search term into the target marketplace's language
3. Choose sort order: If the user wants results sorted by price, reviews, or newness, set the sort parameter
4. Pagination: Use the page parameter to fetch additional result pages if needed
5. Category scope: If the user wants to search within a specific category, provide the node parameter
6. Delivery simulation: Use deliveryZip to see location-specific availability and delivery info

Usage Examples

1. Basic keyword search on US marketplace

{"keyword": "wireless earbuds", "amazonDomain": "amazon.com"}

2. Search on German marketplace with German keyword

{"keyword": "kabellose Kopfhoerer", "amazonDomain": "amazon.de", "language": "de_DE"}

3. Search sorted by price (low to high)

{"keyword": "phone case", "amazonDomain": "amazon.com", "sort": "price-asc-rank"}

4. Search for best sellers in a category

{"keyword": "yoga mat", "amazonDomain": "amazon.com", "sort": "exact-aware-popularity-rank"}

5. Search for newest arrivals on Japan marketplace

{"keyword": "USB充電器", "amazonDomain": "amazon.co.jp", "language": "ja_JP", "sort": "date-desc-rank"}

6. Multi-page search to analyze deeper results

{"keyword": "laptop stand", "amazonDomain": "amazon.com", "page": 2}

7. Mobile device search simulation

{"keyword": "running shoes", "amazonDomain": "amazon.com", "device": "mobile"}

8. Category-scoped search with delivery zip

{"keyword": "office chair", "amazonDomain": "amazon.com", "deliveryZip": "10001"}

Display Rules

1. Present data clearly: Show search results in well-structured tables with key fields: position, ASIN, title, price, rating, review count, brand
2. Highlight sponsored products: Clearly mark which results are sponsored ads vs organic listings
3. Price formatting: Display prices with the correct currency symbol for the marketplace
4. Position context: Remind users that position reflects the actual ranking on the search result page
5. Pagination notice: When results span multiple pages, inform the user how many total results were found and suggest fetching additional pages if needed
6. Error handling: When a query fails, explain the reason based on the error response and suggest adjusting parameters
7. Image links: If image URLs are available, mention them but do not attempt to render them inline unless the user requests it