Skip to main content

xhs-content

XiaoHongShu (小红书) content creation tool. Collects trending XHS notes by keyword, generates a topic research report, then produces XHS-style note drafts. Use when the user wants to: (1) research what's trending on XiaoHongShu/小红书 for a topic, (2) generate XHS note drafts, (3) find XHS content inspiration, or mentions "小红书", "XHS", "xiaohongshu", "xhs content", "xhs note".

Informations de source

Dépôt
mindfold-ai/xhs-content
Dernière activité de la source
11 mars 2026 à 06:49
Langue détectée de SKILL.md
Plusieurs langues
Étoiles
17
Forks
1

Options d'installation

Le prompt qui vérifie d'abord la source est sélectionné par défaut. Vous pouvez passer à une commande directe ou télécharger une copie locale.

Vérifiez les fichiers source

Lisez SKILL.md et les fichiers associés affichés par SkillsMP avant de décider de l'installer.

Affichage de SKILL.md

SKILL.md
Instructions source · Aperçu en lecture seule
name
xhs-content
description
XiaoHongShu (小红书) content creation tool. Collects trending XHS notes by keyword, generates a topic research report, then produces XHS-style note drafts. Use when the user wants to: (1) research what's trending on XiaoHongShu/小红书 for a topic, (2) generate XHS note drafts, (3) find XHS content inspiration, or mentions "小红书", "XHS", "xiaohongshu", "xhs content", "xhs note".
# XHS Content — 小红书选题采集 + 内容生成 从小红书采集热门笔记 → 汇总选题报告 → 生成笔记草稿。 ## Prerequisites - **Python 3** — 运行采集脚本 - **xhs-cli** — 小红书 CLI 工具 (`pipx install xhs-cli` 或 `uv tool install xhs-cli`) - 已登录 (`xhs login` — 支持 QR 码 / cookie / Chrome 自动提取) ## Usage 给定一个关键词,执行三步流程: ### Step 1: 采集 ```bash python3 scripts/xhs-search.py --keyword "<关键词>" --limit 30 ``` 采集结果自动保存到 `data/<date>-<keyword>/sources.jsonl`。 ### Step 2: 汇总报告 读取 `sources.jsonl`,生成选题报告 `data/<date>-<keyword>/report.md`,包含: - 话题热度概览(总笔记数、平均互动) - 热门角度分类(按标题/内容聚类) - Top 笔记摘要(标题、互动数据、核心观点) - 选题建议(基于数据的差异化切入点) - 全部采集笔记表格(标题、作者、赞/藏/评、链接,按互动量降序) ### Step 3: 生成笔记 基于报告 + 固定 prompt,生成 1-3 篇小红书笔记草稿,保存到 `data/<date>-<keyword>/drafts.md`。 ## Workflow (for Claude) When this skill is triggered: 1. Ask user for keyword (if not provided) 2. Run `xhs-search.py` to collect posts → save JSONL 3. Read JSONL, generate report markdown 4. Read report + `references/prompt.md`, generate note drafts 5. Present drafts to user for review ## Data Directory All outputs go to `.claude/skills/xhs-content/data/<date>-<keyword>/`: - `sources.jsonl` — raw collected posts - `report.md` — topic research report - `drafts.md` — generated note drafts ## References - `references/prompt.md` — Fixed generation prompt with XHS style rules
Voir sur GitHub