home / mcp / linkedin mcp server

LinkedIn MCP Server

New Linkedin MCP

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "baptitse-jn-linkedin_mcp": {
      "url": "https://your-site-name.netlify.app/mcp"
    }
  }
}

You run a LinkedIn MCP Server that lets AI assistants interact with LinkedIn data and actions through a unified MCP interface. It includes a Netlify-deployed server for remote usage and a FastAPI client for local testing, with tools to manage profiles, posts, networks, messages, and analytics.

How to use

You access the LinkedIn MCP Server through either a remote deployment or a local setup. Use the MCP client to discover available tools, call them to perform actions, and read resources or analytics provided by the server. Begin by starting the local setup for testing, or access the hosted endpoint to connect your AI assistant environment.

How to install

Prerequisites you need on your machine: Node.js and npm for server scripts, Python for the FastAPI client, and a modern shell. Ensure you have network access to fetch dependencies.

# 1) Clone the project repository
git clone <repository-url>
cd llm_linkedin_mcp_deployment

# 2) Install dependencies for the client and server
npm install

# 3) Move to the MCP client folder and prepare to start
cd mcp-client
""" start_linkedin.sh is used to start the MCP server and the FastAPI client locally """
./start_linkedin.sh

Additional configuration and notes

Remote deployment provides a public MCP endpoint at the designated Netlify URL. Local testing starts the LinkedIn MCP server at http://localhost:8888/mcp and the FastAPI client at http://localhost:8002. When you are ready to test from your development environment, you can run the quick start script and then test with the included helper utilities.

To verify the setup, you can run the quick test script after starting the services. If you need an access token for full LinkedIn functionality, run the OAuth helper to obtain one ready for your client.

Deployment and client usage details

Deploy the MCP server to Netlify by pushing the repository, connecting the project to Netlify, and configuring the build. The published MCP endpoint will be available at the remote /mcp path.

If you use Claude Desktop or another client, configure the MCP server by pointing the client to the Netlify-hosted URL and restart the application to enable MCP integration.

The MCP client exposes an API for interacting with the server. Start the client and access its docs to explore available endpoints for server information, tools, resources, and tool execution.

Troubleshooting and testing tips

Use the MCP Inspector to validate the server locally or at the deployed URL. You can also test with curl against the local endpoint to initialize the MCP, list tools, call tools, and list or read resources.

If something does not respond as expected, check that the local servers are running and that the URL you are using points to the correct mcp path. Review environment variables if you have configured authentication or access controls.

Available tools

profile

Get user profiles and company information to inform AI-driven actions and personalization.

posts

Create and manage LinkedIn posts, including drafting, scheduling, and publishing content.

companies

Search and analyze company profiles to gather intelligence and competitive insights.

connections

Manage connections and send connection requests as part of networking workflows.

messaging

Send and retrieve LinkedIn messages to enable seamless communication flows.

analytics

Provide network analytics and insights to guide engagement strategies.