stickerdaniel/linkedin-mcp-server

What it solves

This project provides a way for AI assistants (such as Claude) to interact with LinkedIn data. It allows these assistants to read profiles, search for jobs and companies, and manage messages, all by leveraging the user's own authenticated browser session rather than relying on a restricted official API.

How it works

It implements the Model Context Protocol (MCP), acting as a server that connects an AI client to a headless Chromium browser. The server can either import an existing LinkedIn session from a locally installed browser (like Chrome or Brave) or prompt the user to log in manually via a browser window. Once authenticated, the the server uses browser automation to scrape and extract data from LinkedIn's web interface.

Who it’s for

Users of MCP-compatible AI agents who want to automate LinkedIn networking, job searching, and lead generation without needing a developer account or complex API keys.

Highlights

• Comprehensive Toolset: Includes tools for fetching person and company profiles, searching for jobs, reading/sending messages, and extracting feed posts.
• Session Import: Ability to automatically import active LinkedIn sessions from common Chromium-based browsers to skip manual login.
• Sesssion Management: Handles browser session persistence and provides serialization to prevent conflicts when multiple MCP clients are running.
• Flexible Deployment: Supports installation via uvx, Docker, or as a Claude Desktop bundle.