home / mcp / linkedin mcp server

LinkedIn MCP Server

A professional MCP server for LinkedIn content creation, publishing, and preview with design-driven components and secure draft management.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "ibm-chuk-mcp-linkedin": {
      "url": "https://linkedin.chukai.io",
      "headers": {
        "OAUTH_ENABLED": "true",
        "SESSION_PROVIDER": "memory",
        "LINKEDIN_CLIENT_ID": "your_linkedin_client_id",
        "LINKEDIN_CLIENT_SECRET": "your_linkedin_client_secret"
      }
    }
  }
}

You can use the LinkedIn MCP Server to compose, preview, publish, and schedule LinkedIn posts with a design-driven component system, secure drafts, and multiple artifact backends. This server enables predictable, data-driven post creation and seamless integration with OAuth flows or direct token usage across desktop and API workflows.

How to use

You will interact with the LinkedIn MCP Server through an MCP client or by running the local server in one of two modes. You can connect a client to a remote MCP URL to handle publishing and drafts, or run a local MCP server and drive it from your development environment.

How to install

Prerequisites you need before installation are Python 3.11 or higher and a LinkedIn OAuth client if you plan to publish.

# Basic installation (STDIO mode only)
pip install chuk-mcp-linkedin

# Recommended: with uv for faster command execution
uv pip install chuk-mcp-linkedin

# Optional extras for HTTP server or document preview
pip install chuk-mcp-linkedin[http]
pip install chuk-mcp-linkedin[preview]

Configuration, usage, and examples

The server supports both a public hosted MCP server and local deployments. You can connect clients to a remote URL or run a local MCP server using a command-line entry point. The following configurations illustrate known MCP usage patterns from practical examples.

{
  "mcpServers": {
    "linkedin": {
      "url": "https://linkedin.chukai.io"
    }
  }
}
```

```json
{
  "mcpServers": {
    "linkedin": {
      "command": "linkedin-mcp",
      "args": ["http", "--port", "8000"],
      "env": {
        "SESSION_PROVIDER": "memory",
        "LINKEDIN_CLIENT_ID": "your_linkedin_client_id",
        "LINKEDIN_CLIENT_SECRET": "your_linkedin_client_secret",
        "OAUTH_ENABLED": "true"
      }
    }
  }
}
```

```json
{
  "mcpServers": {
    "linkedin": {
      "command": "linkedin-mcp",
      "args": ["stdio"]
    }
  }
}

Quick start with a remote MCP server

Connect your MCP client to the remote LinkedIn MCP Server to begin drafting, composing, and previewing posts. You will authenticate and then start drafting content using the provided UI or library.

Available tools

linkedin-mcp

CLI entry point to run the MCP server in various modes (stdio or http) and to manage server configuration.

linkedin_preview_url

Tool that generates session-isolated preview URLs for drafts with configurable expiration.

LinkedInManager

High-level API for managing drafts, previews, and artifact storage across sessions.

ComposablePost

Post composition object that builds text, body, hooks, CTAs, and hashtags from themes.