How an Open‑Source AI Tool Turns Xiaohongshu Posts into Viral Hits

Overview

Xiaohongshu MCP (Multi‑Channel Publisher) is an open‑source AI‑powered tool that automates the end‑to‑end workflow of Xiaohongshu content creation, from topic prediction and copy generation to tag recommendation, data monitoring, competitor analysis, batch account management, and scheduled publishing.

Key Features

AI‑driven topic prediction : Analyzes the last 30 days of Xiaohongshu hot‑list data and user preferences to suggest high‑traffic, low‑competition topics, together with estimated view counts and follower‑growth potential.

Full content generation : Produces three stylistic variants of copy, titles that combine pain points with solutions, and a set of 10+ precise tags in about 30 seconds.

Data monitoring & competitor analysis : Tracks views, likes, saves, and follower gains for each note, visualises trends, and can dissect competitor notes to reveal their topic, structure, and publishing schedule.

Batch account management & scheduled publishing : Supports simultaneous management of five or more accounts, bulk draft saving, one‑click cross‑platform sync (e.g., to Douyin or Weibo), and time‑based posting.

Open‑source and self‑hostable : Source code is publicly available, can be deployed locally for data privacy, and is fully customizable (e.g., custom copy style, tag libraries) without licensing fees.

Typical Use Cases

1. New Blogger – Rapid follower growth

Enter a niche (e.g., "affordable beauty"); the AI predicts a topic such as "Student‑friendly foundation review under $100" with an estimated 100k+ views.

Generate three copy versions, select the most relatable one, and receive ten relevant tags (e.g., #studentbeauty, #budgetfoundation, #oilyskin).

Schedule the post for the AI‑recommended peak hour (e.g., 20:00) and monitor real‑time metrics via the dashboard.

Run competitor analysis on a viral note, extract the winning formula (e.g., "review + budget positioning"), and replicate it for follow‑up notes, each surpassing 50k views.

2. Product‑Focused Blogger – Tripling monetization efficiency

Import a list of products (e.g., ten budget stationery items) in bulk.

AI generates a dedicated title and copy for each item (e.g., "High‑schoolers swear by this mistake‑book – boosted my scores by 50 points!").

The tool auto‑assigns precise tags (e.g., #budgetstationery) and suggests cover images that highlight before‑after score charts.

Schedule two posts per day (09:00 and 20:00), then filter results by "likes‑to‑commission conversion" to prioritize the top three products, raising monthly commission from ¥2,000 to over ¥6,000.

3. Reviving an Old Account – Turning low‑performing notes into hits

Upload a poorly performing note; AI flags three issues: generic title, overly broad tags, and unstructured copy.

Apply AI suggestions: rewrite the title to a more specific one, replace tags with niche alternatives (e.g., #workplacesignals, #newcomerTips, #communication), and restructure the copy into clear bullet points with concrete scenarios.

Republish the optimized note; views jump from ~500 to >80k, likes exceed 20k, and overall account engagement improves dramatically.

Quick Start Guide

Step 1 – Deploy the Tool

Local deployment (recommended for data security) :

# Clone the project
git clone https://github.com/xpzouying/xiaohongshu-mcp.git
cd xiaohongshu-mcp
# Install dependencies
pip install -r requirements.txt
# Copy the example env file and add your Xiaohongshu Cookie
cp .env.example .env
# Launch the application
python main.py

Open a browser and navigate to http://localhost:8501 to access the UI.

Online demo (no installation required) : The README provides a hosted demo link; register and start using core features such as topic prediction and copy generation.

Step 2 – Generate a Note

Click “Create New Note” and fill in the core information: niche (e.g., "affordable beauty"), topic direction (e.g., "foundation review"), and desired style (lifestyle, dry‑facts, or recommendation).

Press “Generate Content”. In about 30 seconds the AI returns a title, body copy, and a set of optimized tags.

Preview the output; if needed, click “Regenerate” or manually tweak details such as adding a personal anecdote.

Step 3 – Publish & Optimize

Schedule the note for an AI‑suggested active hour (e.g., 20:00).

After publishing, monitor the “Data Dashboard” for key metrics: collection rate (indicates content value), comment interaction rate (boosts algorithmic exposure), and follower growth.

Use the “Competitor Analysis” feature to locate similar high‑performing notes, copy their topic angle and publishing time, and continuously refine your own content.

Additional Notes

The project is open source; the community continuously adds features such as AI‑generated cover copy, content‑risk detection, and upcoming fan‑need analysis, aligning the tool with real‑world Xiaohongshu operations.

Project repository: https://github.com/xpzouying/xiaohongshu-mcp