Skip to content
Back to search

gs-skills

SkillWorks with Claude CodeGitHub

Google Scholar skills for Claude Code — search, citation tracking, full-text access, and Zotero export via Chrome DevTools MCP

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/cookjohn--gs-skills/badge.svg)](/listings/cookjohn--gs-skills)

HTML

<a href="/listings/cookjohn--gs-skills"><img src="/api/listings/cookjohn--gs-skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

1. Install Chrome DevTools MCP server

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

2. Install Google Scholar skills

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

Or add to an existing project:

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

3. Start Chrome with remote debugging

# Windows
"C:\Program Files\Google\Chrome\Application\chrome.exe" --remote-debugging-port=9222

# macOS
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222

# Linux
google-chrome --remote-debugging-port=9222

4. Launch Claude Code

claude

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

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 212 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Google Scholar Skills for Claude Code
  2. [English](#english) | [中文](#中文)
  3. | WeChat Official Account (公众号) | WeChat Group (微信群) | Discord |
  4. |:---:|:---:|:---:|
  5. | <img src="qrcode_for_gh_a1c14419b847_258.jpg" width="200"> | <img src="0320.jpg" width="200"> | [Join Discord](https://discord.gg/tGd5vTDASg) |
  6. | 未来论文实验室 | 扫码加入交流群 | English & Chinese |
  7. ---
  8. <a id="english"></a>
  9. ## English
  10. [Claude Code](https://docs.anthropic.com/en/docs/claude-code) skills that let Claude interact with [Google Scholar](https://scholar.google.com) through Chrome DevTools MCP.
  11. Search papers, track citations, get full-text links, and export to Zotero — all from the Claude Code CLI.
  12. ### Prerequisites
  13. - [Claude Code](https://docs.anthropic.com/en/docs/claude-code) CLI installed
  14. - Chrome browser with remote debugging enabled
  15. - [Zotero](https://www.zotero.org/) desktop app (optional, for export)
  16. - Python 3 (optional, for Zotero push script)
  17. ### Skills
  18. | Skill | Description | Invocation |
  19. |-------|-------------|------------|
  20. | `gs-search` | Keyword search with structured result extraction | `/gs-search deep learning` |
  21. | `gs-advanced-search` | Filtered search: author, journal, date range, exact phrase, title-only | `/gs-advanced-search author:Einstein after:2020 relativity` |
  22. | `gs-cited-by` | Find papers that cite a given paper via data-cid | `/gs-cited-by 0qfs6zbVakoJ` |
  23. | `gs-fulltext` | Get full-text access links: PDF, DOI, Sci-Hub, publisher | `/gs-fulltext 0qfs6zbVakoJ` |
  24. | `gs-navigate-pages` | Pagination for search results | `/gs-navigate-pages next` |
  25. | `gs-export` | Export to Zotero via BibTeX extraction | `/gs-export 0qfs6zbVakoJ` |
  26. ### Agent
  27. **`gs-researcher`** — orchestrates all 6 skills. Handles CAPTCHA detection (pauses and asks user to solve manually) and multi-step workflows like "search → track citations → export to Zotero".
  28. ### Installation
  29. #### 1. Install Chrome DevTools MCP server
  30. ```bash
  31. claude mcp add chrome-devtools -- npx -y chrome-devtools-mcp@latest
  32. ```
  33. #### 2. Install Google Scholar skills
  34. ```bash
  35. git clone https://github.com/cookjohn/gs-skills.git
  36. cd gs-skills
  37. cp -r skills/ agents/ .claude/
  38. ```
  39. Or add to an existing project:
  40. ```bash
  41. git clone https://github.com/cookjohn/gs-skills.git /tmp/gs-skills
  42. cp -r /tmp/gs-skills/skills/ your-project/.claude/skills/
  43. cp -r /tmp/gs-skills/agents/ your-project/.claude/agents/
  44. ```
  45. #### 3. Start Chrome with remote debugging
  46. ```bash
  47. # Windows
  48. "C:\Program Files\Google\Chrome\Application\chrome.exe" --remote-debugging-port=9222
  49. # macOS
  50. /Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222
  51. # Linux
  52. google-chrome --remote-debugging-port=9222
  53. ```
  54. #### 4. Launch Claude Code
  55. ```bash
  56. claude
  57. ```
  58. Skills and agent are picked up automatically. Try `/gs-search deep learning` to verify.
  59. ### How It Works
  60. All skills use async `evaluate_script` calls via Chrome DevTools MCP — no screenshot parsing or OCR. Each skill operates in 1-2 tool calls (navigate + evaluate_script), making interactions fast and reliable.
  61. Key design choices:
  62. - **DOM scraping only** — Google Scholar has no public API; all data extraction uses CSS selectors
  63. - **`data-cid` as primary key** — cluster ID used across all skills for citation tracking, export, and cross-referencing
  64. - **Single async script per operation** — replaces multi-step snapshot → click → wait_for patterns
  65. - **BibTeX via navigate_page** — bypasses CORS restrictions on `scholar.googleusercontent.com`
  66. - **CAPTCHA-aware** — detects Google Scholar CAPTCHA and pauses for manual resolution
  67. ### Project Structure
  68. ```
  69. skills/
  70. ├── gs-search/SKILL.md # Basic keyword search
  71. ├── gs-advanced-search/SKILL.md # Filtered search (author, journal, date, etc.)
  72. ├── gs-cited-by/SKILL.md # Citation tracking via data-cid
  73. ├── gs-fulltext/SKILL.md # Full-text access links (PDF, DOI, Sci-Hub)
  74. ├── gs-navigate-pages/SKILL.md # Result pagination
  75. ├── gs-export/ # BibTeX export & Zotero push
  76. │ ├── SKILL.md
  77. │ └── scripts/
  78. │ └── push_to_zotero.py # Zotero Connector API client
  79. agents/
  80. └── gs-researcher.md # Agent: orchestrates all skills
  81. ```
  82. ---
  83. <a id="中文"></a>
  84. ## 中文
  85. | 公众号 | 微信交流群 | Discord |
  86. |:---:|:---:|:---:|
  87. | <img src="qrcode_for_gh_a1c14419b847_258.jpg" width="200"> | <img src="0320.jpg" width="200"> | [加入 Discord](https://discord.gg/tGd5vTDASg) |
  88. | 未来论文实验室 | 扫码加入交流群 | 中英文交流 |
  89. 让 [Claude Code](https://docs.anthropic.com/en/docs/claude-code) 通过 Chrome DevTools MCP 操控 [Google Scholar (谷歌学术)](https://scholar.google.com) 的技能集。
  90. 支持论文搜索、引用追踪、全文获取、导出到 Zotero 等功能,全部在 Claude Code 命令行中完成。
  91. ### 前置要求
  92. - 已安装 [Claude Code](https://docs.anthropic.com/en/docs/claude-code) CLI
  93. - Chrome 浏览器(需开启远程调试)
  94. - [Zotero](https://www.zotero.org/) 桌面版(可选,用于导出)
  95. - Python 3(可选,用于 Zotero 推送脚本)
  96. ### 技能列表
  97. | 技能 | 描述 | 调用方式 |
  98. |------|------|----------|
  99. | `gs-search` | 关键词搜索,返回结构化结果 | `/gs-search deep learning` |
  100. | `gs-advanced-search` | 高级搜索:作者、期刊、时间、精确短语、仅标题 | `/gs-advanced-search author:Einstein after:2020 relativity` |
  101. | `gs-cited-by` | 引用追踪:查找引用了某篇论文的所有文献 | `/gs-cited-by 0qfs6zbVakoJ` |
  102. | `gs-fulltext` | 全文获取:PDF、DOI、Sci-Hub、出版商链接 | `/gs-fulltext 0qfs6zbVakoJ` |
  103. | `gs-navigate-pages` | 搜索结果翻页 | `/gs-navigate-pages next` |
  104. | `gs-export` | 通过 BibTeX 导出到 Zotero | `/gs-export 0qfs6zbVakoJ` |
  105. ### 智能体
  106. **`gs-researcher`** — 统一调度全部 6 个技能。自动检测验证码(暂停并提示用户手动完成),支持"搜索 → 引用追踪 → 导出到 Zotero"等复合工作流。
  107. ### 安装步骤
  108. #### 1. 安装 Chrome DevTools MCP 服务器
  109. ```bash
  110. claude mcp add chrome-devtools -- npx -y chrome-devtools-mcp@latest
  111. ```
  112. #### 2. 安装 Google Scholar 技能
  113. ```bash
  114. git clone https://github.com/cookjohn/gs-skills.git
  115. cd gs-skills
  116. cp -r skills/ agents/ .claude/
  117. ```
  118. 或添加到已有项目:
  119. ```bash
  120. git clone https://github.com/cookjohn/gs-skills.git /tmp/gs-skills
  121. cp -r /tmp/gs-skills/skills/ your-project/.claude/skills/
  122. cp -r /tmp/gs-skills/agents/ your-project/.claude/agents/
  123. ```
  124. #### 3. 启动 Chrome 远程调试
  125. ```bash
  126. # Windows
  127. "C:\Program Files\Google\Chrome\Application\chrome.exe" --remote-debugging-port=9222
  128. # macOS
  129. /Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222
  130. # Linux
  131. google-chrome --remote-debugging-port=9222
  132. ```
  133. #### 4. 启动 Claude Code
  134. ```bash
  135. claude
  136. ```
  137. 技能和智能体会自动加载。输入 `/gs-search deep learning` 验证是否正常工作。
  138. ### 工作原理
  139. 所有技能通过 Chrome DevTools MCP 的 `evaluate_script` 异步执行 JavaScript,无需截图识别或 OCR。每个技能仅需 1-2 次工具调用(导航 + 执行脚本),快速且稳定。
  140. 核心设计:
  141. - **纯 DOM 解析** — Google Scholar 无公开 API,所有数据通过 CSS 选择器提取
  142. - **`data-cid` 作为主键** — 集群 ID 贯穿所有技能,用于引用追踪、导出和交叉引用
  143. - **单次异步脚本** — 取代多步骤 snapshot → click → wait_for 模式
  144. - **navigate_page 获取 BibTeX** — 绕过 `scholar.googleusercontent.com` 的 CORS 限制
  145. - **验证码感知** — 检测到 Google Scholar 验证码时自动暂停,等待用户手动完成
  146. ---
  147. ## License
  148. MIT

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.