home / mcp / linkedin mcp server

LinkedIn MCP Server

Provides MCP endpoints to query LinkedIn data such as profiles, companies, and job information through various transport options.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "stickerdaniel-linkedin-mcp-server": {
      "command": "uvx",
      "args": [
        "linkedin-scraper-mcp"
      ],
      "env": {
        "HOST": "127.0.0.1",
        "PORT": "8000",
        "TIMEOUT": "10000",
        "CHROME_PATH": "/usr/bin/chromium"
      }
    }
  }
}

You can connect AI assistants to LinkedIn data through this MCP server, enabling you to access profiles, companies, job postings, and job details via a consistent programmatic interface. It supports multiple ways to run and connect, including local shells, Docker containers, Claude Desktop integration, and development setups.

How to use

To use this MCP server from your MCP client, start the server using one of the supported runtimes, then point your client to the local MCP endpoint. The server exposes a set of tools that let you fetch person and company profiles, search for jobs and people, and retrieve job details. You can run in standard local mode for quick testing or run inside Docker for containerized deployments. If you choose the HTTP transport option, you enable web-based MCP clients to connect by running the server with the streamable HTTP transport and the appropriate host/port/path.

How to install

Choose your preferred installation method and follow the steps below. The server supports local development, Docker deployment, and integration with Claude Desktop.

Using the Universal UVX setup (recommended)

Prerequisites: Install uv and then set up a persistent browser profile for LinkedIn access.

Step-by-step UVX setup and run

uvx linkedin-scraper-mcp --login
```
This opens a browser for you to log in manually. The browser profile is saved to your home directory under the LinkedIn MCP profile path.

uvx linkedin-scraper-mcp --login
```
This opens a browser for you to log in manually. The browser profile is saved to `~/.linkedin-mcp/profile/`.

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

Starting the server with UVX (normal run)

To start the LinkedIn MCP server locally, run the following command. This starts the server in standard stdio mode for local MCP clients.

Starting the server with UVX (debug/login flow)

To create or refresh the login session, run the login command again when needed.

HTTP mode (web-based MCP clients) with UVX

If you want to expose the server to web-based MCP clients, you can run with the transport set to streamable-http and specify host, port, and path.

Client configuration example (UVX-based runtime)

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

Available tools

get_person_profile

Fetch person profile data with explicit sections such as experience, education, interests, honors, languages, contact_info, and posts.

get_company_profile

Extract company profile data with explicit sections like posts and jobs.

get_company_posts

Retrieve recent posts from a company LinkedIn feed.

search_jobs

Search LinkedIn jobs using keywords and location filters.

search_people

Search LinkedIn profiles by keywords and location.

get_job_details

Get detailed information about a specific job posting.

close_session

Close the browser session and clean up resources.