🤖 Agent‐Native Skills – 源自工作流的实操沉淀
Practical tools derived from real‐world workflows
WorkBuddy Compatible OpenClaw Compatible ClawHub License: MIT PRs Welcome GitHub Stars
🚀 Pull Requests Welcome | 💬 Issues Welcome | 🌟 Star Us | 📦 Available on ClawHub
This repository collects a few agent‐native skills derived from my daily workflows.
Each skill targets a narrow, authentic problem that waste handsome time:
- Scattered documentation sites that need to be merged into a single PDF for offline reading.
- Academic data collection blocked by basic anti‐bot measures (IP‐based rate limiting).
If you face similar annoyances, these tools might save you a few hours. Constructive criticism and better ideas are always welcome.
| Skill | Description | Use Case | Difficulty |
|---|---|---|---|
| gitbook-to-pdf | Convert GitBook documentation to PDF with recursive crawling | Documentation archival, offline reading | ⭐⭐ Easy |
| anti-bot-downloader | Academic research web scraper with IP rotation | Data collection for research | ⭐⭐⭐ Medium |
- Python 3.8+
- Node.js 18+ (for md-to-pdf)
- pip3
# Clone the repository git clone https://github.com/RecelFrone/Andy-agent-skill-workshop.git cd Andy-agent-skill-workshop # Install all dependencies chmod +x install.sh ./install.sh # Or install manually: # 1. Install npm packages npm install -g md-to-pdf # 2. Install Python packages pip3 install -r skills/anti-bot-downloader/requirements.txt
gitbook-to-pdf:
# Use with WorkBuddy or OpenClaw # Just say: "Convert https://docs.example.com to PDF" # Or use the script directly: python3 skills/gitbook-to-pdf/scripts/gitbook_crawler.py \ --base-url "https://docs.example.com" \ --start-path "/docs/" \ --output-dir "/tmp/output"
anti-bot-downloader:
# For academic research data collection python3 skills/anti-bot-downloader/scripts/downloader.py \ --url "https://example.com/data" \ --output ~/Downloads/
| Name | Description | Status |
|---|---|---|
| gitbook-to-pdf | GitBook documentation to PDF converter | ✅ Stable |
| anti-bot-downloader | Anti-crawler web downloader for research | ✅ Stable |
Bugs or suggestions → Issues Direct improvements → Pull Requests
Again, Honest reflection is much appreciated!!!
MIT – free to use, attribution appreciated.
这个仓库记录了我从日常工作中拆出的几个 agent‐native skill。 每个 skill 都对应一类反复出现、边界清晰的小麻烦:
- 文档站点结构零散,想一次性抓下来合并成 PDF 离线阅读。
- 学术数据采集被基础反爬策略(IP 限频)卡住,手动换 IP 太累。 如果你也踩过类似的坑,这些脚本或许能帮上忙。欢迎任何批评和指教!
| Skill | 描述 | 使用场景 | 难度 |
|---|---|---|---|
| gitbook-to-pdf | GitBook 文档转 PDF,支持递归抓取 | 文档归档、离线阅读 | ⭐⭐ 简单 |
| anti-bot-downloader | 学术研究用反爬虫下载器 | 科研数据采集 | ⭐⭐⭐ 中等 |
- Python 3.8+
- Node.js 18+ (用于 md-to-pdf)
- pip3
# 克隆仓库 git clone https://github.com/RecelFrone/Andy-agent-skill-workshop.git cd Andy-agent-skill-workshop # 一键安装所有依赖 chmod +x install.sh ./install.sh # 或者手动安装: # 1. 安装 npm 包 npm install -g md-to-pdf # 2. 安装 Python 包 pip3 install -r skills/anti-bot-downloader/requirements.txt
gitbook-to-pdf 场景:
💬 "帮我把这个 GitBook 文档转成 PDF" - 你只需要说一句话,AI 就能帮你完成整个网站的抓取和 PDF 生成!
# 使用 WorkBuddy 或 OpenClaw # 只需要说:"把 https://docs.example.com 转成 PDF"
anti-bot-downloader 场景:
💬 "帮我下载这个学术网站的数据做研究" - AI 会自动处理反爬机制,帮你安全高效地采集数据!
# 用于学术研究数据采集 python3 skills/anti-bot-downloader/scripts/downloader.py \ --url "https://example.com/data" \ --output ~/Downloads/
| 名称 | 描述 | 状态 |
|---|---|---|
| gitbook-to-pdf | GitBook 文档转 PDF 转换器 | ✅ 稳定版 |
| anti-bot-downloader | 学术研究反爬虫下载器 | ✅ 稳定版 |
问题报告或改进建议 → Issues 代码贡献 → Pull Requests
再三说明,非常欢迎各位大佬拍砖指教!
MIT 协议 — 可自由使用,保留署名即可。
Made with ❤️ by the community, for the community
用 ❤️ 由社区打造,为社区服务