Unlinked

Bring your LinkedIn professional profile into AI assistants — without copy-pasting, scraping, or guessing.

Unlinked is a TypeScript MCP server that connects your LinkedIn professional profile and experience to any Model Context Protocol client — Claude Desktop, GitHub Copilot, and others. It reads your data directly from LinkedIn's official Member Data Portability API, so your assistant always has accurate, up-to-date context about who you are professionally.

:eu: EEA / Switzerland only. LinkedIn's Member Data Portability API is currently available exclusively to members located in the European Economic Area and Switzerland. Thank you, Digital Markets Act (DMA)!

What it does

Once connected, your AI assistant can:

• Read your full professional profile — positions, education, skills, certifications, projects, languages, honours, publications, patents, and more
• Fetch any specific LinkedIn data domain by name
• Poll for recent profile changes via the LinkedIn changelog API
• Diagnose authorization status so you can quickly identify consent or permission issues

All data access is read-only. Your access token is used only to talk to LinkedIn and is never stored, logged, or echoed back.

Prerequisites

• Node.js 22+
• A LinkedIn Developer app with the Member Data Portability API (Member) product enabled (follow this guide)
• A member access token from the LinkedIn OAuth Token Generator (requires EEA/Switzerland membership)

Setup

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "unlinked": {
      "command": "npx",
      "args": ["-y", "@larsbaunwall/unlinked"],
      "env": {
        "LINKEDIN_TOKEN": "<your_access_token>"
      }
    }
  }
}

The config file is typically at:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

GitHub Copilot (VS Code)

Add to your user-level MCP config (File → Preferences → MCP Servers) or a workspace .vscode/mcp.json:

{
  "servers": {
    "unlinked": {
      "command": "npx",
      "args": ["-y", "@larsbaunwall/unlinked"],
      "env": {
        "LINKEDIN_TOKEN": "<your_access_token>"
      }
    }
  }
}

Tools

Configuration

Development

npm install
npm run build
npm run dev        # run directly from source with .env
npm run inspect    # test with MCP Inspector

See AGENTS.md for implementation guidance.

License

MIT