← 返回 Skills 市场

微信公众号文章抓取 (Jeff版)

Name: 微信公众号文章抓取 (Jeff版)
Author: jeffli2002

作者 jeffli2002 · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ✓ 安全检测通过

146

总下载

当前安装

版本数

在 OpenClaw 中安装

/install jeffli-wechat-article

功能描述

微信公众号文章抓取工具。将微信公众号文章转换为 Markdown 格式，支持图片本地下载。当用户提到抓取微信公众号文章、提取公众号内容、爬取微信文章时触发。

使用说明 (SKILL.md)

微信公众号文章抓取

将微信公众号文章转换为 Markdown 格式，支持图片本地下载。

脚本位置

主程序：scripts/main.py
MCP Server：scripts/mcp_server.py

快速使用

cd ~/.openclaw/workspace/skills/wechat-article
python3 scripts/main.py "文章URL" -o /root/.openclaw/workspace/output

参数

参数	说明
`-o DIR`	输出目录（默认 ./output）
`-v`	调试日志
`--no-images`	不下载图片，保持远程 URL
`--force`	覆盖已存在文件
`--no-headless`	显示浏览器（用于处理验证码）

输出结构

output/
 └── 文章标题/
     ├── 文章标题.md
     └── images/
         ├── img_001.jpg
         └── ...

注意事项

验证码：遇到验证页面时加 --no-headless 手动处理
反爬：微信有频率限制，建议间隔操作
图片失败：保留远程 URL，可用 --force 重试

依赖

camoufox
markdownify
beautifulsoup4
httpx
aiohttp

安全使用建议

This skill appears to do what it says: fetch WeChat article pages in a headless browser, convert HTML to Markdown, and download images into an output folder. Before installing or running: - Be aware it performs live web requests and will download a camoufox browser binary on first run — run it in an isolated environment if you want to limit risk. - You may need to pip-install requirements (requirements.txt) before running; confirm dependency sources. - Running the MCP server exposes conversion tools to any MCP-capable client (over stdio by default); that lets an AI invoke scraping and write files — only enable the server for trusted clients. - If you need stricter assurance, review the camoufox package and its download behavior (where the browser binary comes from) and run the tool manually on a sample URL in a safe workspace first.

功能分析

Type: OpenClaw Skill Name: jeffli-wechat-article Version: 1.0.0 The skill is a well-structured and legitimate tool for converting WeChat Official Account articles into Markdown format. It uses the Camoufox stealth browser (scripts/scraper.py) to handle WeChat's anti-bot measures and JavaScript rendering, and includes robust logic for concurrent image downloading (scripts/downloader.py) and metadata extraction (scripts/parser.py). No evidence of data exfiltration, malicious execution, or prompt injection was found; the code logic is entirely consistent with its stated purpose.

能力评估

✓ Purpose & Capability

Name/description match the code and scripts: the package fetches rendered WeChat pages (camoufox), parses content, converts to Markdown, and optionally downloads images. The use of a stealth browser (camoufox) and an MCP server is coherent for anti-detection scraping and AI integration.

✓ Instruction Scope

SKILL.md instructs the agent to run the included CLI or MCP server and to write outputs into an output directory. The runtime instructions and code operate on article URLs and local output paths only; they do not ask the agent to read unrelated files, environment variables, or exfiltrate data to unexpected endpoints.

ℹ Install Mechanism

The registry contains no formal install spec (instruction-only), but the repository includes requirements.txt and uses camoufox which will auto-download a browser binary on first run. Running the skill will therefore pull packages (pip) and possibly large browser artifacts from the internet — review and run in a controlled environment if you do not trust those origins.

✓ Credentials

The skill declares no required environment variables or credentials and the code does not access secrets. Network access is required to fetch WeChat pages and images (expected for a scraper). There are minor dependency-list mismatches (SKILL.md lists aiohttp while requirements.txt lists mcp), but nothing requesting unrelated credentials.

✓ Persistence & Privilege

The skill does not request always:true and does not modify other skills or system-wide configs. It exposes an MCP server (stdio transport) so an AI client can invoke its tools; autonomous invocation is the platform default and not a concern by itself. Outputs are written to the provided output directory (and debug HTML when parsing fails).

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install jeffli-wechat-article
安装完成后，直接呼叫该 Skill 的名称或使用 /jeffli-wechat-article 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

版本历史

v1.0.0

首次发布：微信公众号文章抓取工具，支持转换为Markdown格式和图片本地下载

元数据

Slug jeffli-wechat-article

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题