Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Wechat Xhs Publisher

v1.0.0

微信公众号与小红书一键发布工具。当需要将热点新闻改写成公众号文章并发布到微信公众号和小红书时使用,包括:1)根据热点新闻改写公众号文章 2)使用AI生成文章配图 3)发布前IP检测 4)发布到微信公众号 5)发布到小红书。触发场景:热点新闻发布、公众号文章发布、小红书种草、社交媒体同步发布。

0· 216·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for achievejia/wechat-xhs-publisher.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Wechat Xhs Publisher" (achievejia/wechat-xhs-publisher) from ClawHub.
Skill page: https://clawhub.ai/achievejia/wechat-xhs-publisher
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install wechat-xhs-publisher

ClawHub CLI

Package manager switcher

npx clawhub@latest install wechat-xhs-publisher
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The skill claims to publish to WeChat and Xiaohongshu and the instructions do that, but it delegates publishing to external tools (baoyu-post-to-wechat, mcporter) and an IP-checker skill. The registry metadata declares no required binaries or credentials, yet the runtime commands (npx, bun, mcporter, mcporter call, image API) clearly require specific tools and likely account credentials. The absence of declared requirements is an incoherence.
Instruction Scope
Instructions are focused on the stated workflow: generate article, generate images, check public IP, then publish. They reference network endpoints (ip38.com and image APIs) and call other local/third-party tools. They instruct reading/writing workspace files (post-to-wechat/YYYY-MM-DD/, img/, {workspace}/wechat_ip_config.md) which is within the task's scope. However, they also instruct running npx -y bun scripts/wechat-api.ts (which will execute downloaded code) and calling mcporter — both of which may perform arbitrary network I/O and require credentials not declared here.
!
Install Mechanism
There is no install spec (instruction-only), which is lower risk in principle, but the runtime relies on commands that will fetch/run code at execution (npx -y bun ...). Using npx -y invokes package download/execution from registries at runtime, which can execute arbitrary code. The skill does not declare required binaries (node/npx/bun/mcporter), so the metadata is inconsistent with the commanded runtime environment.
!
Credentials
The skill declares no required environment variables or primary credential, yet publishing to WeChat and Xiaohongshu will require authentication tokens/credentials handled by the referenced tools. That mismatch leaves it unclear how credentials are supplied and audited. The skill also reads a workspace config file for an IP whitelist — verify this file doesn't store other secrets. Network endpoints and image-generation APIs may require API keys not declared here.
Persistence & Privilege
The skill does not request always:true, does not modify other skills, and is instruction-only. It writes files into the workspace and invokes other tools, which is expected for a publishing workflow. Autonomous invocation is allowed by default but is not in itself a new risk here.
What to consider before installing
Before installing or using this skill: 1) Recognize the metadata is incomplete — the SKILL.md expects node/npx/bun and mcporter and will run npx -y which downloads and executes packages at runtime (this can run arbitrary code). 2) Confirm where and how WeChat/Xiaohongshu credentials will be provided and whether the referenced tools (baoyu-post-to-wechat, mcporter, wechat-ip-checker) are trusted and already configured. 3) Inspect the workspace file {workspace}/wechat_ip_config.md and ensure it contains only the IP record (no tokens/credentials). 4) If you must test, do so in an isolated environment or sandbox account to avoid accidental public posting. 5) Prefer explicit declaration of required binaries and environment variables from the skill author, and ask for documentation on the external tools it invokes and how they handle authentication and network traffic.

Like a lobster shell, security has layers — review code before you run it.

latestvk97461ncjs2dxndzya8e30snbx8349bd
216downloads
0stars
1versions
Updated 5h ago
v1.0.0
MIT-0

微信公众号与小红书发布工具

将热点新闻一键改写并发布到微信公众号和小红书。

工作流程

[热点新闻] → [Step1:公众号文章创作] → [Step2:AI配图生成] → [Step3:IP检测] → [Step4:公众号发布] → [Step5:小红书发布]

Step 1: 公众号文章创作

输入

  • 热点新闻素材(标题、内容要点)
  • 写作风格要求

核心命令

根据热点新闻和写作风格,生成一篇新的公众号文章:

  1. 分析热点新闻的核心要点
  2. 按照指定风格创作文章
  3. 生成Markdown格式文章,包含frontmatter元数据

写作风格

  • 根据用户指定风格执行
  • frontmatter中的author使用用户提供的作者名

输出

  • Markdown格式文章
  • 保存在 post-to-wechat/YYYY-MM-DD/ 目录
  • frontmatter包含:title, author, date, source, cover_image

Step 2: AI配图生成

核心命令

根据文章内容,分析需要生成的配图和插图数量,调用图像API生成对应图片:

  1. 分析文章结构和内容要点
  2. 确定需要几张配图(封面+内文插图)
  3. 为每张图片生成合适的prompt
  4. 调用图像API生成图片

配图规划

根据文章内容确定配图数量和风格

工具

使用 baoyu-image-gen 或直接调用图像API

保存路径

  • 复制到文章目录:post-to-wechat/YYYY-MM-DD/img/
  • 并保存到工作区:img/

Step 3: IP检测

核心命令

在发布公众号文章前,调用 wechat-ip-checker 技能检测公网IP:

  1. 获取当前公网IP(通过 ip38.com)
  2. 与配置文件中的记录IP比对
  3. IP无变化 → 继续执行公众号发布
  4. IP有变化 → 中断流程,通知用户确认

工具

使用 wechat-ip-checker skill

配置文件

  • 位置:{workspace}/wechat_ip_config.md
  • 格式:
# 微信公众号发布公网IP记录
公网IP地址: xxx.xxx.xxx.xxx
IP查询来源: https://www.ip38.com/

判断逻辑

  • ✅ IP相同:继续执行 Step 4
  • ❌ IP不同:中断流程,提示用户添加新IP到白名单

Step 4: 公众号发布

工具

使用 baoyu-post-to-wechat

发布命令

cd <baoyu-post-to-wechat路径>
npx -y bun scripts/wechat-api.ts <文章md文件> \
  --theme default \
  --cover <封面图绝对路径> \
  --author "<作者名称>"

关键参数

  • --theme default 必填
  • --cover 使用绝对路径
  • 配图路径需转换为相对路径(相对于文章md文件)

重要

  • 发布时勾选原创声明
  • 确保图片与文章一起上传

Step 5: 小红书发布

工具

使用 xiaohongshu MCP (mcporter)

发布命令

mcporter call 'xiaohongshu.publish_content(
  title: "标题",
  content: "正文",
  images: ["图片绝对路径1", "图片绝对路径2"],
  tags: ["话题1", "话题2"],
  is_original: true
)'

关键参数

  • is_original: true 必须勾选原创
  • 图片使用绝对路径
  • 标题≤20字
  • 正文不要包含#标签(通过tags传递)

常见问题

Q: 图片路径问题

A: 封面图使用绝对路径;文章内配图使用相对于md文件的相对路径,并将图片复制到md文件同目录下

Q: 小红书漏掉原创

A: 必须在参数中添加 is_original: true

Q: 公众号图片未上传

A: 确保图片在md文件同目录下,脚本会自动上传

输出报告格式

## ✅ 发布完成

### 公众号
- 状态:草稿已保存
- media_id:xxx

### 小红书
- 状态:发布完成
- 原创:✅

Comments

Loading comments...