Token导航 LogoToken导航TokenDH.com
研究检索操作浏览器github未标认证来源可访问许可证需确认审计异常

linkedinLinkedIn 职场运营

Agent Skill

linkedin 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

233

周安装

10

GitHub Stars

公开资料未说明

下载量

82
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:linkedin(LinkedIn 职场运营)
来源仓库:https://github.com/skillhq/linkedin
仓库路径:skills/linkedin
安装命令:
npx skills add https://github.com/skillhq/linkedin --skill linkedin
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/skillhq/linkedin --skill linkedin

简介

用于查找、检索和筛选相关信息。

  • 适合在 LinkedIn 职场运营相关场景中使用。
  • 可通过来源仓库和 README 核验具体用法。
  • 安装命令:npx skills add https://github.com/skillhq/linkedin --skill linkedin。
  • 建议确认权限范围及是否会触发联网或文件读写。

SKILL.md

LinkedIn Skill

Access LinkedIn data through browser automation. This approach is reliable because it uses your actual browser session, avoiding LinkedIn's aggressive bot detection that blocks API-only traffic.

Setup Requirements

Chrome Profile Setup (One-Time)

  1. Create a dedicated Chrome profile for automation:

- Open Chrome → click your profile icon (top right) → "Add" - Name it something memorable (e.g., "Claude" or "Automation") - This keeps automation separate from personal browsing

  1. In your automation profile:

- Log into LinkedIn (www.linkedin.com) - Install the Claude browser extension from https://claude.ai/chrome - Keep this Chrome window open when using LinkedIn commands

  1. Before using LinkedIn commands:

- Ensure Chrome is open with your automation profile - The Claude browser extension must be connected

Browser Automation Commands

When the user asks for LinkedIn data, use the claude-in-chrome tools to navigate and extract information.

Connecting to Chrome

Always start by checking the browser connection:

1. Call mcp__claude-in-chrome__tabs_context_mcp with createIfEmpty: true
2. If not connected, ask user to open Chrome with their automation profile
3. Create a new tab or use existing one

Reading Feed

When user asks to see their LinkedIn feed:

1. Navigate to https://www.linkedin.com/feed/
2. Use read_page to extract feed content
3. Parse and format the feed items

Example flow:

  1. mcp__claude-in-chrome__navigate to https://www.linkedin.com/feed/
  2. mcp__claude-in-chrome__read_page with filter: "all" to get post content
  3. Extract post authors, content, engagement metrics from the accessibility tree

Viewing Profiles

When user asks about a LinkedIn profile:

1. Navigate to https://www.linkedin.com/in/{username}/
2. Use read_page to extract profile information
3. Parse name, headline, about, experience, etc.

Searching

When user wants to search LinkedIn:

People:    https://www.linkedin.com/search/results/people/?keywords={query}
Jobs:      https://www.linkedin.com/search/results/jobs/?keywords={query}
Companies: https://www.linkedin.com/search/results/companies/?keywords={query}
Posts:     https://www.linkedin.com/search/results/content/?keywords={query}

Reading Messages

When user wants to check messages:

1. Navigate to https://www.linkedin.com/messaging/
2. Use read_page to extract conversation list
3. Click on a conversation to read messages

Data Extraction Patterns

Feed Posts

Look for these patterns in the accessibility tree:

  • Post author: Links with profile URLs
  • Post content: Text nodes within post containers
  • Engagement: Like/comment/share counts
  • Timestamps: Relative time indicators

Profile Data

Navigate to profile and extract:

  • Name and headline from main header
  • About section text
  • Experience entries (company, title, duration)
  • Education entries
  • Skills section

Search Results

Parse result items containing:

  • Name/title link
  • Subtitle (headline, location)
  • Connection degree
  • Profile picture

Example Interactions

"Show my LinkedIn feed"

1. Connect to Chrome: tabs_context_mcp
2. Navigate: navigate to linkedin.com/feed/
3. Wait for load: computer action=wait duration=2
4. Read content: read_page with depth=10
5. Parse and summarize top posts

"Look up John Smith on LinkedIn"

1. Connect to Chrome: tabs_context_mcp
2. Navigate: navigate to linkedin.com/search/results/people/?keywords=John%20Smith
3. Read results: read_page filter=interactive
4. Present matching profiles

"What's my profile looking like?"

1. Connect to Chrome: tabs_context_mcp
2. Navigate: navigate to linkedin.com/in/me/
3. Read profile: read_page
4. Summarize profile stats, recent activity

CLI Fallback (Less Reliable)

The CLI tool is available but LinkedIn often invalidates API sessions:

# Install
cd /Users/derekrein/Code/cyberdrk305/linkedin && npm install && npm run build && npm link

# Commands (may require fresh cookies frequently)
linkedin setup              # Setup instructions
linkedin check              # Verify credentials
linkedin whoami             # Show logged-in user
linkedin feed -n 10         # Get feed posts
linkedin profile <username> # View profile

Manual Cookie Setup

If using the CLI, you'll need to manually extract cookies:

  1. Open Chrome DevTools (F12) on linkedin.com
  2. Go to: Application → Cookies → linkedin.com
  3. Copy li_at and JSESSIONID values
  4. Run: linkedin setup --li-at "VALUE" --jsessionid "VALUE"

Note: LinkedIn aggressively invalidates these sessions after a few API calls.

Troubleshooting

Browser extension not connected

  • Ensure Chrome is open with your automation profile
  • Click the Claude extension icon to activate it
  • Try restarting Chrome

LinkedIn checkpoint/verification page

  • LinkedIn may require verification for new browser sessions
  • Complete the verification manually, then retry

Session expired

  • Log back into LinkedIn in your automation Chrome profile
  • Refresh the page and try again

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

38.1%
按下载量换算31

Claude

29.1%
按下载量换算24

Cursor

19.32%
按下载量换算16

Gemini CLI

9.53%
按下载量换算8

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

未通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills