---
title: "LinkedIn MCP Server for Saved Posts"
description: "LinkedIn ships no MCP server. LinkedMash runs a hosted saved posts MCP server at mcp.linkedmash.com that works with Claude, ChatGPT, Cursor and any MCP client."
canonical: "https://www.linkedmash.com/linkedin-mcp-server"
last-updated: "2026-08-10"
---

# LinkedIn MCP Server for Saved Posts

LinkedIn ships no MCP server. LinkedMash runs a hosted saved posts MCP server at mcp.linkedmash.com that works with Claude, ChatGPT, Cursor and any MCP client.

Canonical HTML: https://www.linkedmash.com/linkedin-mcp-server
Agent-readable Markdown: https://www.linkedmash.com/linkedin-mcp-server.md

Last updated: 2026-08-10

Part of [For AI agents](https://www.linkedmash.com/for-ai-agents).

## Summary

**Is there a LinkedIn MCP server?**

Not from LinkedIn. LinkedIn publishes no Model Context Protocol server and no saved-posts API, so an agent cannot reach your saves on its own. LinkedMash runs a hosted saved posts MCP server at mcp.linkedmash.com: the extension imports your LinkedIn saved posts once, then any MCP client can search, label, draft and schedule against your own library.

## How it works

### 1. Import your saved posts

Add the LinkedMash extension. It imports your existing LinkedIn saved posts with the post text, author and media intact.

### 2. Generate an API token

In LinkedMash, open Settings then Developer API and generate a token. It authenticates every MCP and REST call.

### 3. Point your client at the server

Add https://mcp.linkedmash.com/api/mcp?token=YOUR_TOKEN as an MCP server in your client config, restart it, and ask for your recent saved posts.

### No server to install

The MCP server is hosted, so there is nothing to clone, build or keep running. One URL and a token connect any client.

- One hosted endpoint: mcp.linkedmash.com/api/mcp
- Token auth from Settings, Developer API
- Works with any spec-compliant MCP client

### What the agent can actually do

The server exposes your library as tools, not as a blob of text, so the agent queries it the way you would.

- Search saved posts by keyword or meaning
- Read a single post, count the library, list labels
- Add and remove labels in bulk
- Draft, schedule and analyze posts on Premium

### MCP for agents, REST for code

The same library is available over a documented REST API, so your own scripts and automations use the identical data.

- Bearer-token REST endpoints with docs
- Webhooks for save, label and sync events
- Trigger CSV, JSON and PDF exports programmatically

## Frequently asked questions

### Does LinkedIn have an official MCP server?

No. LinkedIn publishes no Model Context Protocol server, and its public APIs cover marketing, talent and share use cases without exposing your saved posts. Any MCP access to a LinkedIn library today comes from a third-party tool that captures the posts from your own browser session first.

### What is the LinkedMash MCP server URL?

The hosted server is https://mcp.linkedmash.com/api/mcp. Clients authenticate with an API token generated in LinkedMash under Settings then Developer API, passed as a token query parameter in the server URL.

### Which clients can connect to it?

Any client that speaks MCP. That includes Claude and Claude Code, ChatGPT, Cursor, VS Code, Windsurf, Codex and Raycast. They all use the same hosted server URL, so there is no per-client build.

### Do I need to self-host anything?

No. The server is hosted by LinkedMash, so there is no repository to clone and no process to keep alive. You add one URL to your client config and the connection works from any machine that has your token.

### Which plan includes MCP access?

You can try MCP and the Developer API on the free plan with a 20-result sample and 100 calls per UTC day. A paid plan unlocks the full library, paging, filters and higher limits. Premium also includes drafting, scheduling and analytics over the same connection. Lifetime holders keep API and MCP access as part of the plan.

### Can the agent post to LinkedIn for me?

Yes, on Premium. The scheduling tools put a draft into your queue and publish through LinkedIn official Posts API at the slot you picked, using the LinkedIn account you connected. Nothing automates the LinkedIn interface in your browser.

## Get started

**Connect the MCP server**

Import your saves, generate a token, and point your client at mcp.linkedmash.com.

[Read the MCP docs](https://www.linkedmash.com/login)

## Import your library

[Start with LinkedMash](https://www.linkedmash.com/?utm_source=agent_markdown&utm_medium=ai&utm_campaign=public_markdown&utm_content=linkedin-mcp-server), then [install the extension and import your LinkedIn saved posts](https://chromewebstore.google.com/detail/linkedmash/ldadjiiepooimpnjagjidpklcpbnamop).