linkedin-post in your AI assistant

A free MCP server, no account needed: your assistant writes the post, our tools format it for LinkedIn with correct Unicode characters, accents included.

Server URLhttps://linkedin-post.com/mcp

Available tools

format_linkedin_post
Turns **bold**, *italic*, __underline__ and ~~strikethrough~~ into text ready to paste on LinkedIn, with the exact character count.
unformat_linkedin_text
Converts a post copied from LinkedIn back to plain text and editable markers.
analyze_linkedin_post
Checks length, hook, paragraphs, bold, hashtags and shows what is visible before “see more”.
add_breathing_room
Adds breathing room to dense text without changing a word: isolated hook, short blocks, lists on their own lines.

Setup

Claude (claude.ai and app)

Settings → Connectors → Add custom connector.
Name: linkedin-post. URL: https://linkedin-post.com/mcp
Enable the connector in a conversation.

Claude Code

In a terminal:

claude mcp add --transport http linkedin-post https://linkedin-post.com/mcp

ChatGPT

Settings → Apps & Connectors → Advanced settings → enable developer mode.
Create a connector with the URL https://linkedin-post.com/mcp, no authentication.

Cursor, Windsurf, VS Code

Add to your MCP config file (for example ~/.cursor/mcp.json):

{
  "mcpServers": {
    "linkedin-post": { "url": "https://linkedin-post.com/mcp" }
  }
}

Gemini CLI

Add to ~/.gemini/settings.json:

{
  "mcpServers": {
    "linkedin-post": { "httpUrl": "https://linkedin-post.com/mcp" }
  }
}

Codex CLI and other clients

For stdio-only clients, use the mcp-remote bridge (example for ~/.codex/config.toml):

[mcp_servers.linkedin-post]
command = "npx"
args = ["-y", "mcp-remote", "https://linkedin-post.com/mcp"]

Example prompt

Write a LinkedIn post about my 3 freelancing lessons, bold the key ideas with linkedin-post, then analyze the hook.

Tools run on our server with no AI and no storage: text is processed then forgotten. Limit of 60 calls per minute.