Xiaohongshu Note Search

What Xiaohongshu Note Search does

Xiaohongshu Note Search is a Node.js command-line skill that returns structured data on public Xiaohongshu notes. Four scripts cover keyword search, note details, comment retrieval and creator monitoring, with filters for content type, sort order, time range and result count. A single run can fetch up to 10,000 items. Results come back as standard JSON, and each run is archived in a local log.

The tool does not log in to a Xiaohongshu account and reads only public content. It offers no write actions such as posting, liking or following. It has no dependencies beyond Node.js. The data comes from a paid API service at guaikei.com, so a TOKEN set as GUAIKEI_API_TOKEN is required. The repository ships a SKILL.md and is tagged for several agent platforms, including Hermes. The README is written in Chinese.

Key features

• Keyword search for notes with likes, collects and comments
• Note detail retrieval with text, images and engagement data
• Comment retrieval for sentiment and opinion analysis
• Creator monitoring, including follower, like and collect totals
• Filters for content type, sort order, time range and count
• JSON output and a SKILL.md for agent use

When to use it

• Finding the highest-liked notes for a keyword to plan content topics
• Pulling the comment section of a note for opinion analysis
• Checking a competitor or creator account's recent posts

Who it is for: Xiaohongshu creators, marketers and analysts who run AI agents, including Hermes Agent, for content research.

How it fits with Hermes Agent

The repository is tagged hermes-skill and hermes-skills among many agent platforms. The README names WorkBuddy, OpenClaw, QClaw, Claude Code and Cursor as agents and does not describe Hermes-specific steps.

Requirements: Node.js and a paid API TOKEN from guaikei.com, set as GUAIKEI_API_TOKEN

Note: Data comes from a paid third-party API at guaikei.com that requires a TOKEN.

FAQ