Publora MCP server

Publora MCP Server

Official Model Context Protocol (MCP) server for Publora — control your social media scheduling directly from AI assistants like Claude, Cursor, and any MCP-compatible client.

No coding required. Just describe what you want in plain English:

"Schedule a LinkedIn post for tomorrow at 9am"

"How did my last post perform?"

"Post this to Twitter and LinkedIn"

Quick Start

Remote Server (Recommended)

Publora hosts an MCP server at mcp.publora.com/mcp — no installation needed.

Using Claude? Install it from the connectors directory: claude.ai/directory/publora. Select Connect, sign in to Publora, approve. No API key, no config file — on web, desktop and mobile alike.

Every other client:

1. Get your API key at app.publora.com/dashboard/api (API in the dashboard sidebar)

2. Add to your MCP client:

{
  "mcpServers": {
    "publora": {
      "type": "http",
      "url": "https://mcp.publora.com/mcp",
      "headers": {
        "Authorization": "Bearer sk_YOUR_API_KEY"
      }
    }
  }
}

Clients that support OAuth (Claude Code, Cursor, VS Code, Codex and others) can leave out headers and sign in through the browser instead.

3. Restart your client and start talking to your AI about social media!

Client Setup

Claude Code (CLI)

claude mcp add --transport http --scope user publora https://mcp.publora.com/mcp

Run /mcp, select publora, choose Authenticate and sign in to Publora in the browser. No API key needed.

For CI or headless machines, send a static key instead:

claude mcp add --transport http --scope user publora https://mcp.publora.com/mcp \
  --header "Authorization: Bearer sk_YOUR_API_KEY"

The same entry can go in ~/.claude.json or a project .mcp.json (JSON above). Verify with the /mcp command.

Claude Desktop

Claude Desktop uses the same connectors as Claude on the web: open claude.ai/directory/publora and select Connect. There is no config file to edit.

Settings → Developer → Edit Config is only needed to run this repository locally over stdio, after cloning it and running npm install (see Local Development for how its tools differ from the hosted server):

{
  "mcpServers": {
    "publora": {
      "command": "node",
      "args": ["/path/to/mcp-server/src/stdio.js"],
      "env": {
        "PUBLORA_API_KEY": "sk_YOUR_API_KEY"
      }
    }
  }
}

Cursor

Create .cursor/mcp.json in your project:

{
  "mcpServers": {
    "publora": {
      "type": "http",
      "url": "https://mcp.publora.com/mcp",
      "headers": {
        "Authorization": "Bearer sk_YOUR_API_KEY"
      }
    }
  }
}

Local Development

Run your own instance:

git clone https://github.com/publora/mcp-server.git
cd mcp-server
npm install
PUBLORA_API_URL=https://api.publora.com npm start

Server runs on http://localhost:3100

The code in this repository is an April 2026 snapshot of the server. Its tool set is older: it has LinkedIn analytics and workspace tools, and lacks completemedia, deletemedia, prunemediareference, poststats, profilestats, linkedincreatereshare and linkedinlistmentionables. Use the hosted server for the current tools.

Available Tools (18)

The hosted server at mcp.publora.com/mcp exposes:

Shortened. The full README is on GitHub.