n8n workflow

LinkedIn MCP Server

Manage LinkedIn posts and engagement with an MCP server and AI agents.

By Tushar Rayamajhi · 15 nodes · 14 tools

The canvas loads after the page. You can also open it on n8n.io.

Open on n8n.io

What it does

This workflow exposes a LinkedIn toolkit through an MCP server, so an MCP-compatible agent (Claude, Cursor or an n8n AI Agent) can act on your LinkedIn account on request.

The agent can publish text, image and article-link posts, read your profile and recent posts, edit or delete a post, and comment on, react to or remove engagement on your own content.

How it works

An MCP Server Trigger hosts the tools. You do not run the workflow by hand: the agent calls a tool and n8n runs the matching LinkedIn node.

The canvas groups the tools into profile access, author feed and text posts, post lookup and deletion, publish and engage, and update and moderate.

Every node reads a single linkedin_person_urn Variable, so there is nothing to find and replace across the workflow.

Destructive tools (delete post, delete comment, remove reaction) are there when you want them. Add your own safeguards before you give an agent access.

Tools and integrations

  • MCP Server Trigger (Bearer auth)
  • LinkedIn node and LinkedIn REST API through HTTP Request tool nodes
  • LinkedIn OAuth2
  • 14 tools: posts, comments, reactions, profile
  • Works with Claude, Cursor or an n8n AI Agent

Setup and requirements

You bring your own accounts and credentials. Nothing in the JSON contains a secret: keys live in n8n credentials, Variables or the app's Settings page.

  1. Create a LinkedIn Developer app with the Share on LinkedIn and Sign In with LinkedIn using OpenID Connect products, then a LinkedIn OAuth2 credential in n8n.
  2. Create a Bearer Auth credential for the MCP Server Trigger and reuse the same token in your AI client.
  3. In Settings, Variables, add linkedin_person_urn. Run the Fetch LinkedIn Profile tool once to find your ID.
  4. Copy the MCP Server Trigger production URL into your MCP client.
  5. For image posts, send the image as binary data and pass the right binary field name. Post analytics and company-page posting need LinkedIn Marketing Developer Platform approval.

Full step-by-step notes are on the canvas itself as sticky notes, and on the original n8n.io template.

Want a workflow like this built for your business, or have an idea? Message us

Tell us what your team repeats every day. We design and build n8n workflows and AI agents, with a fixed price and a written timeline.

Message ushello@solvex.com.np