LinkedIn MCP Server

Read LinkedIn profiles, companies and job posts with your own login.

Started in our test
★ 4K GitHub stars
49 downloads a week
Last commit 26 Oct 2026

At a glance

This server uses a browser session signed in to your LinkedIn account to read profiles, company pages and job listings, and to search jobs. It is a scraper, not an official LinkedIn API.

That has two consequences: it can break when LinkedIn changes its pages, and heavy use can get an account restricted. Use it for occasional lookups.

Made by: Daniel Sticker (an independent project, not made by LinkedIn)
Licence: Apache-2.0
Latest version: 4.14.0 (10 Jun 2026)
PyPI package: linkedin-scraper-mcp
GitHub stars: 4K
Downloads: 49 a week
Last commit: 26 Oct 2026
Tools: 20
Runs on: Your computer

Our check: We ran this exact config on 26 Oct 2026: it started and answered with 20 tools. The npm package "linkedin-mcp-server" has no link to this project, and the PyPI package of that name belongs to someone else. This project publishes to PyPI as linkedin-scraper-mcp.

What's in the pod: its tools

We found 20 tools when we started it on 26 Oct 2026. 17 only read. The rest can change things, so keep your app's "ask before each tool call" setting on for them.

Changes things3 tools · Can create, edit, send or delete.

Tool	What it is for
`close_session`	Shuts the browser it uses for LinkedIn and frees the profile.
`connect_with_person`	Sends a connection request, with a note if you want one.
`send_message`	Sends a LinkedIn message to someone.

Only reads17 tools · Looks things up and changes nothing.

Tool	What it is for
`get_company_employees`	Lists people who work at a company.
`get_company_posts`	Reads a company page's recent posts.
`get_company_profile`	Reads a company page: what it does, size, location and more.
`get_conversation`	Opens one message thread and reads it.
`get_feed`	Reads posts from your LinkedIn home feed.
`get_inbox`	Lists your recent message threads.
`get_job_apply_url`	Finds the link to apply for a job.
`get_job_details`	Reads a job post in full.
`get_my_profile`	Reads your own profile.
`get_person_profile`	Reads someone's profile: experience, education, skills and more.
`get_saved_jobs`	Lists jobs you have saved.
`get_sidebar_profiles`	Reads the people LinkedIn suggests alongside a profile.
`search_companies`	Searches LinkedIn for companies.
`search_conversations`	Searches your messages by keyword.
`search_jobs`	Searches job listings with filters such as location and date.
`search_people`	Searches LinkedIn for people by keywords and filters.
`search_posts`	Searches posts by keyword, optionally from the last day, week or month.

Grouped and described by us. The README on GitHub lists the inputs each tool takes.

Set it up

Pick your app for the exact snippet and where it goes.

Where it goes: claude_desktop_config.json (Settings → Developer → Edit Config)

Claude Desktop

{
  "mcpServers": {
    "linkedin": {
      "command": "uvx",
      "args": [
        "linkedin-scraper-mcp"
      ]
    }
  }
}

Where it goes: Your terminal, in the project folder

Claude Code

claude mcp add-json linkedin '{"command":"uvx","args":["linkedin-scraper-mcp"]}'

Where it goes: ~/.cursor/mcp.json, or .cursor/mcp.json in a project

Cursor

{
  "mcpServers": {
    "linkedin": {
      "command": "uvx",
      "args": [
        "linkedin-scraper-mcp"
      ]
    }
  }
}

Where it goes: .vscode/mcp.json in your project

VS Code

{
  "servers": {
    "linkedin": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "linkedin-scraper-mcp"
      ]
    }
  }
}

Where it goes: ~/.codeium/windsurf/mcp_config.json

Windsurf

{
  "mcpServers": {
    "linkedin": {
      "command": "uvx",
      "args": [
        "linkedin-scraper-mcp"
      ]
    }
  }
}

New to this? See the setup guides for Claude Code, Cursor, Claude Desktop and ChatGPT.

Settings

Names are exactly as the project spells them. Required ones are marked.

Setting	What it does
`--login`Command line flag	Opens a browser so you can sign in to LinkedIn once; the session is saved for later.`uvx mcp-server-linkedin@latest --login`
`--import-from-browser`Command line flag	Reuses the LinkedIn sign-in from Chrome, Brave, Edge, Arc and other Chromium browsers instead of logging in again.`--import-from-browser brave`
`--logout`Command line flag	Deletes the saved LinkedIn session.
`--timeout`Command line flag	Milliseconds to wait for each page step. Defaults to 5000. Also TIMEOUT.`--timeout 10000`
`--tool-timeout`Command line flag	Seconds a whole tool call may take. Defaults to 180. Also TOOL_TIMEOUT.`--tool-timeout 300`
`--no-headless`Command line flag	Shows the browser window so you can watch what it does.
`--proxy-server`Command line flag	Routes the browser through a proxy. Set it before you log in.`--proxy-server http://host:port`
`--transport`Command line flag	stdio by default, or streamable-http to run it as a local web server.

What it can touch

It can read

Anything your LinkedIn account can see: profiles, companies, jobs, your feed and your messages
Cookies from a local browser if you import a session

It can change

Sends messages and connection requests as you
Stores your LinkedIn session in ~/.linkedin-mcp on your computer

Safer setup: Use it for reading and research, and approve every send_message and connect_with_person call by hand. Heavy automated use can get a LinkedIn account restricted.

Before you start

A LinkedIn account
uv (for uvx)
A one-time sign-in (the project README shows the current login step)

If it does not work

LinkedIn shows a captcha or asks for app confirmation: Run with --login so a real browser opens, solve the check yourself, and confirm in the LinkedIn mobile app if asked.
Elements not found or pages hang: Raise --timeout to 10000 or more.
Whole tool calls time out: Raise --tool-timeout, for example to 300, especially for long profiles or on a cold start.
Browser import fails on newer Chrome: Chrome 127 and later encrypts cookies so they cannot always be read. Use --login instead.

Peas in a pod: similar servers

How LinkedIn MCP Server lines up against other knowledge and work apps servers on the shelf.

Server	Tools	Runs on	Licence	Stars
LinkedIn MCP Server	20	Your computer	Apache-2.0	4K
arXiv MCP Server	19	Your computer	Apache-2.0	3.2K
Notion	24	Your computer or hosted	MIT	4.7K
MCP Atlassian (Jira and Confluence)	98	Your computer	–	6K
Basic Memory	27	Your computer	AGPL-3.0-or-later	4.4K