CNKI Skills for Claude Code

English | 中文

WeChat Official Account (公众号)	WeChat Group (微信群)	Discord
		Join Discord
未来论文实验室	扫码加入交流群	English & Chinese

English

Claude Code skills that let Claude interact with CNKI (中国知网) through Chrome DevTools MCP.

Search papers, browse journals, check indexing status, download PDFs, and export to Zotero — all from the Claude Code CLI.

Prerequisites

Claude Code CLI installed
Chrome browser (login handled manually for downloads)
Zotero desktop app (optional, for export)
Python 3 (optional, for Zotero push script)

Skills

Skill	Description	Invocation
`cnki-search`	Keyword search with structured result extraction	`/cnki-search 人工智能`
`cnki-advanced-search`	Filtered search: author, journal, date, source category (SCI/EI/CSSCI/北大核心/CSCD)	`/cnki-advanced-search CSSCI 人工智能 2020-2025`
`cnki-parse-results`	Re-parse an existing search results page	`/cnki-parse-results`
`cnki-navigate-pages`	Pagination and sort order	`/cnki-navigate-pages next`
`cnki-paper-detail`	Extract full paper metadata (abstract, keywords, etc.)	`/cnki-paper-detail`
`cnki-journal-search`	Find journals by name, ISSN, or CN	`/cnki-journal-search 计算机学报`
`cnki-journal-index`	Check journal indexing status and impact factors	`/cnki-journal-index 计算机学报`
`cnki-journal-toc`	Browse issue table of contents	`/cnki-journal-toc 计算机学报 2025 01期`
`cnki-download`	Download paper PDF/CAJ	`/cnki-download`
`cnki-export`	Push to Zotero or output GB/T 7714 citation	`/cnki-export zotero`

Agent

cnki-researcher — orchestrates all 10 skills. Handles captcha detection (pauses and asks user to solve manually), tab management, and multi-step workflows like "search → filter → export to Zotero".

Installation

1. Install Chrome DevTools MCP server

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

2. Install CNKI skills

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

Or add to an existing project:

git clone https://github.com/cookjohn/cnki-skills.git /tmp/cnki-skills
cp -r /tmp/cnki-skills/skills/ your-project/.claude/skills/
cp -r /tmp/cnki-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 /cnki-search 深度学习 to verify.

How It Works

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.

Key design choices:

Single async script per operation — replaces multi-step snapshot → click → wait_for patterns
navigate_page over link clicking — CNKI opens links in new tabs; navigating directly avoids tab management overhead
Batch export — export multiple papers from search results in one call instead of visiting each detail page
Captcha-aware — detects Tencent slider captcha and pauses for manual resolution

Project Structure

skills/
├── cnki-search/SKILL.md            # Basic keyword search
├── cnki-advanced-search/SKILL.md   # Filtered search (source category, date, etc.)
├── cnki-parse-results/SKILL.md     # Parse existing results page
├── cnki-navigate-pages/SKILL.md    # Pagination & sorting
├── cnki-paper-detail/SKILL.md      # Paper metadata extraction
├── cnki-journal-search/SKILL.md    # Journal lookup
├── cnki-journal-index/SKILL.md     # Journal indexing & impact factors
├── cnki-journal-toc/SKILL.md       # Issue table of contents
├── cnki-download/SKILL.md          # PDF/CAJ download
└── cnki-export/                    # Citation export & Zotero
    ├── SKILL.md
    └── scripts/
        └── push_to_zotero.py       # Zotero Connector API client
agents/
└── cnki-researcher.md              # Agent: orchestrates all skills

中文

公众号	微信交流群	Discord
		加入 Discord
未来论文实验室	扫码加入交流群	中英文交流

让 Claude Code 通过 Chrome DevTools MCP 操作中国知网 (CNKI) 的技能集。

支持论文检索、期刊浏览、收录查询、PDF 下载、导出到 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（下载功能需手动登录知网）
Zotero 桌面端（可选，用于导出）
Python 3（可选，用于 Zotero 推送脚本）

技能列表

技能	功能	调用方式
`cnki-search`	关键词检索，返回结构化结果	`/cnki-search 人工智能`
`cnki-advanced-search`	高级检索：作者、期刊、时间、来源类别（SCI/EI/CSSCI/北大核心/CSCD）	`/cnki-advanced-search CSSCI 人工智能 2020-2025`
`cnki-parse-results`	重新解析当前搜索结果页	`/cnki-parse-results`
`cnki-navigate-pages`	翻页与排序	`/cnki-navigate-pages next`
`cnki-paper-detail`	提取论文详细信息（摘要、关键词等）	`/cnki-paper-detail`
`cnki-journal-search`	按名称、ISSN、CN 号查找期刊	`/cnki-journal-search 计算机学报`
`cnki-journal-index`	查询期刊收录情况和影响因子	`/cnki-journal-index 计算机学报`
`cnki-journal-toc`	浏览期刊目录	`/cnki-journal-toc 计算机学报 2025 01期`
`cnki-download`	下载论文 PDF/CAJ	`/cnki-download`
`cnki-export`	推送到 Zotero 或输出 GB/T 7714 引用	`/cnki-export zotero`

智能体

cnki-researcher — 统一调度全部 10 个技能。自动处理验证码检测（暂停并提示用户手动完成）、标签页管理，支持"检索 → 筛选 → 导出到 Zotero"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 CNKI 技能

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

添加到已有项目：

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

3. 启动 Chrome 远程调试

# 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. 启动 Claude Code

claude

技能和智能体会自动加载。输入 /cnki-search 深度学习 验证是否正常。

工作原理

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。每个操作仅需 1-2 次工具调用（导航 + 执行脚本），快速且稳定。

核心设计：

单次异步脚本 — 取代多步骤的 snapshot → click → wait_for 模式
直接导航优于点击链接 — 知网链接会打开新标签页，直接导航避免标签页管理开销
批量导出 — 从搜索结果页一次性导出多篇论文，无需逐篇进入详情页
验证码感知 — 检测到腾讯滑块验证码时自动暂停，等待用户手动处理

License

MIT

cnki-skills

Links

README