/install xiaohongshu-search-mcp
XiaoHongShu Search MCP
Use this skill when the user wants to search XiaoHongShu / RED / RedNote and receive structured JSON results through the hosted whatson.red MCP endpoint.
Requirements
- The user needs a whatson.red account with an Agent API key from https://www.whatson.red/agents.
- Read the API key from
WHATSON_RED_API_KEY. - Send the key as
Authorization: Bearer ${WHATSON_RED_API_KEY}. - MCP requests use account-based whatson.red billing and credits.
MCP Endpoint
Use the Streamable HTTP endpoint:
https://www.whatson.red/api/agent/mcp
Primary tool:
search_xiaohongshu_notes
Tool Arguments
querystring, required: Search query in English, Chinese, or pinyin.limitnumber, optional: 1-100, default 60. Values above 20 fetch additional search pages.include_note_detailsboolean, optional: Default true. Includes note detail payloads when available.include_blogboolean, optional: Default false. Returns a draft blog/article payload without publishing it.
Use limit values such as 40, 60, 80, or 100 when the user needs more than one page of notes. Note details are best-effort for the returned real notes.
Expected Response
The tool returns structured content with:
querytranslatedQuerysearchQuerynotesnoteDetailsblogblogErrorusage
usage includes charged credits, remaining credits, and whether the request was already charged through idempotency.
REST Alternative
For non-MCP callers, whatson.red also exposes:
POST https://www.whatson.red/api/agent/search
REST supports API-key billing and accountless x402 payment. x402 is REST-only; do not use x402 for MCP. REST x402 searches cost $0.05 USDC per raw search on supported USDC networks, and includeBlog=true is not available through x402.
Safety Notes
- Do not ask the user for Browserless, proxy, cookie, or deployment secrets; this skill only uses the hosted whatson.red MCP endpoint.
- Do not attempt to scrape XiaoHongShu directly from this skill.
- Do not publish posts when
include_blogis true; the returned blog payload is only a draft.
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install xiaohongshu-search-mcp - 安装完成后,直接呼叫该 Skill 的名称或使用
/xiaohongshu-search-mcp触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
XiaoHongShu Search MCP 是什么?
Search XiaoHongShu / RED notes through the hosted whatson.red MCP endpoint when users need structured RED search results, note details, or a draft blog payload. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 106 次。
如何安装 XiaoHongShu Search MCP?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install xiaohongshu-search-mcp」即可一键安装,无需额外配置。
XiaoHongShu Search MCP 是免费的吗?
是的,XiaoHongShu Search MCP 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
XiaoHongShu Search MCP 支持哪些平台?
XiaoHongShu Search MCP 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 XiaoHongShu Search MCP?
由 Chris(@seichris)开发并维护,当前版本 v1.0.0。