Sd Skills

ScienceDirect Skills for Claude Code

English | 中文

English

Claude Code skills that let Claude interact with ScienceDirect (Elsevier) through Chrome DevTools MCP.

Search papers, browse journals, extract metadata, export citations, download PDFs, and push to Zotero — all from the Claude Code CLI.

Prerequisites

• Claude Code CLI installed
• Chrome browser (institutional or personal login for PDF download & export)
• Zotero desktop app (optional, for citation export)
• Python 3 (optional, for Zotero push script)

Skills

Agent

sd-researcher — orchestrates all 8 skills. Handles Cloudflare captcha detection (auto-clicks or pauses for manual resolution), and supports multi-step workflows like "search → detail → export to Zotero → download PDF".

Installation

1. Install Chrome DevTools MCP server

claude mcp add chrome-devtools -- npx -y chrome-devtools-mcp@latest

2. Install ScienceDirect skills

git clone https://github.com/cookjohn/sd-skills.git
cd sd-skills
cp -r skills/ agents/ .claude/

Or add to an existing project:

git clone https://github.com/cookjohn/sd-skills.git /tmp/sd-skills
cp -r /tmp/sd-skills/skills/ your-project/.claude/skills/
cp -r /tmp/sd-skills/agents/ your-project/.claude/agents/

3. Configure anti-detection (recommended)

Add to your .claude.json to prevent Cloudflare captcha loops on PDF downloads:

{
  "mcpServers": {
    "chrome-devtools": {
      "command": "npx",
      "args": [
        "-y",
        "chrome-devtools-mcp@latest",
        "--ignoreDefaultChromeArg=--enable-automation",
        "--ignoreDefaultChromeArg=--disable-infobars",
        "--chromeArg=--disable-blink-features=AutomationControlled"
      ]
    }
  }
}

4. Launch Claude Code

claude

Skills and agent are picked up automatically. Try /sd-search deep learning to verify.

Project Structure

skills/
├── sd-search/SKILL.md              # Basic keyword search
├── sd-advanced-search/SKILL.md     # Filtered search (author, journal, year, etc.)
├── sd-parse-results/SKILL.md       # Parse existing results page
├── sd-navigate-pages/SKILL.md      # Pagination & sorting
├── sd-paper-detail/SKILL.md        # Paper metadata extraction
├── sd-journal-browse/SKILL.md      # Journal info & issue browsing
├── sd-download/SKILL.md            # PDF download with Cloudflare handling
└── sd-export/                      # Citation export & Zotero
    ├── SKILL.md
    └── scripts/
        └── push_to_zotero.py       # Zotero Connector API client
agents/
└── sd-researcher.md                # Agent: orchestrates all skills

中文

让 Claude Code 通过 Chrome DevTools MCP 操作 ScienceDirect (Elsevier) 的技能集。

支持论文检索、期刊浏览、元数据提取、引用导出、PDF 下载、推送到 Zotero 等功能，全部在 Claude Code 命令行中完成。

前置要求

• 已安装 Claude Code CLI
• Chrome 浏览器（下载和导出需机构或个人账号登录）
• Zotero 桌面端（可选，用于引用导出）
• Python 3（可选，用于 Zotero 推送脚本）

技能列表

智能体

sd-researcher — 统一调度全部 8 个技能。自动处理 Cloudflare 验证码（自动点击或暂停等待用户手动完成），支持"检索 → 详情 → 导出到 Zotero → 下载 PDF"等复合工作流。

安装方法

1. 安装 Chrome DevTools MCP 服务器

claude mcp add chrome-devtools -- npx -y chrome-devtools-mcp@latest

2. 安装 ScienceDirect 技能

git clone https://github.com/cookjohn/sd-skills.git
cd sd-skills
cp -r skills/ agents/ .claude/

添加到已有项目：

git clone https://github.com/cookjohn/sd-skills.git /tmp/sd-skills
cp -r /tmp/sd-skills/skills/ your-project/.claude/skills/
cp -r /tmp/sd-skills/agents/ your-project/.claude/agents/

3. Configure anti-detection（推荐）

在 .claude.json 中添加以下配置，优化 PDF 下载体验：

{
  "mcpServers": {
    "chrome-devtools": {
      "command": "npx",
      "args": [
        "-y",
        "chrome-devtools-mcp@latest",
        "--ignoreDefaultChromeArg=--enable-automation",
        "--ignoreDefaultChromeArg=--disable-infobars",
        "--chromeArg=--disable-blink-features=AutomationControlled"
      ]
    }
  }
}

4. 启动 Claude Code

claude

技能和智能体会自动加载。输入 /sd-search deep learning 验证是否正常。

License

MIT