Documentation

PostSpread MCP Server

The official Model Context Protocol (MCP) server for PostSpread. Empower AI agents to seamlessly interact with your PostSpread account .

Introduction

This package empowers AI agents (like Claude Desktop, Cursor, and custom bots) to seamlessly interact with your PostSpread account. Your AI can read your analytics, list your connected social media accounts, upload media files, and publish or schedule cross-platform posts natively.

Setup takes less than 60 seconds for most platforms!

Getting Started

Quick Setup

Choose your AI platform below and follow the setup instructions:

Claude (Desktop & Web)

Requirements:

Setup Steps:

  1. 1 In Claude, click the Settings icon (⚙️)
  2. 2 Go to Connectors → Custom connectors
  3. 3 Click Add custom connector
  4. 4 Enter the connector URL:
https://mcp.postspread.com/mcp/sse
  1. 5 Click Connect — a browser window will open for OAuth authentication
  2. 6 Log in to your PostSpread account and click Allow to authorize

Pro tip:

Once connected, just ask Claude naturally:

Schedule a post about our new product for tomorrow at 3pm across Instagram and Facebook

Or:

Show me last week's engagement metrics for all my posts

ChatGPT

Requirements:

Step 1: Enable Developer Mode

  1. 1 Open ChatGPT
  2. 2 Click Settings (gear icon)
  3. 3 Go to Developer tab
  4. 4 Turn Developer Mode on

Step 2: Connect PostSpread

  1. 1 Go to ChatGPT Plugins
  2. 2 Click the + (New Plugin) button
  3. 3 Fill in the plugin details:
  • Name: PostSpread
  • Description: Manage and schedule social media posts across all platforms
  • Connection: Select "Server URL"
  • Server URL: https://mcp.postspread.com/mcp/sse
  • Authentication: Select "OAuth"
  1. 4 Check "I understand and want to continue"
  2. 5 Click Create
  3. 6 You'll be redirected to log in to your PostSpread account
  4. 7 Click Allow to authorize ChatGPT to access your PostSpread account

Step 3: Start Using

Once connected, start a new chat and type:

@PostSpread  Show my connected social accounts.

Or ask naturally:

Schedule a post about our new product for tomorrow at 3pm on Instagram and Facebook

Grok

Requirements:

Setup Steps:

  1. 1 Go to grok.com/connectors
  2. 2 Click New Connector and select "Custom"
  3. 3 Give it a preferred name: PostSpread
  4. 4 Enter the connector URL:
https://mcp.postspread.com/mcp/sse
  1. 5 Click Connect — a browser window will open for OAuth sign-in
  2. 6 Log in to your PostSpread account and click Allow to grant permissions

Tip: Grok only requests the permissions it needs. You can manage or revoke access anytime from your Grok settings.

Claude Code (CLI)

Requirements:

Setup Steps:

  1. 1 Open your terminal and navigate to your Claude Code project directory
  2. 2 Create or edit your Claude Code configuration file (.claude/config.json or via Claude Code settings)
  3. 3 Add the PostSpread MCP server to your configuration:
{
  "mcpServers": {
    "postspread": {
      "command": "npx",
      "args": ["-y", "postspread-mcp"],
      "env": {
        "POSTSPREAD_API_TOKEN": "your-postspread-api-token-here"
      }
    }
  }
}
  1. 4 Restart Claude Code CLI or reload your configuration

Example Usage:

Create a post announcing our new feature release and schedule it for tomorrow at 2pm on Twitter and LinkedIn

Note: Find your API token in PostSpread → Settings → Integrations → API Access.

Code Editors

Connecting AI Code Editors

Modern AI code editors like Cursor, Windsurf, and Zed natively support local MCP servers via stdio.

Setup Steps:

  1. 1 Open your editor's Settings (e.g. Cursor Settings → Features → MCP)
  2. 2 Click + Add New MCP Server
  3. 3 Set the type to command
  4. 4 Name it postspread and set the command to:
npx -y postspread-mcp
  1. 5 You will be prompted to enter your environment variables. Add your POSTSPREAD_API_TOKEN
Advanced

Connecting Custom Web Agents

If you are building your own custom AI agent (using LangChain, OpenClaw, Vercel AI, etc.) and want it to post to social media automatically, you can run this package as a hosted Server-Sent Events (SSE) server!

Setup Steps:

  1. 1 Clone this repository and deploy it to a Node.js server (e.g. Heroku, Vercel, Railway)
  2. 2 Start the server using node server.js (Defaults to Port 2407)
  3. 3 Point your @modelcontextprotocol/sdk client to http://your-server.com/mcp/sse
  4. 4 Pass your user's PostSpread API Token in the HTTP Authorization headers when establishing the SSE connection
API Reference

Available AI Tools

Once connected, your AI agent automatically gains access to a robust suite of tools:

Posts & Publishing

create_post

Create and schedule a post. Use is_draft to save without publishing. Pass the social account IDs from list_social_accounts — the correct workspace is resolved automatically. Supports platform_configurations to override platform-specific settings.

list_posts

List your posts with optional filters by platform or status.

get_post

Get full details of a single post by ID.

update_post

Update a scheduled or draft post's caption, schedule, accounts, or media. Pass the social account IDs from list_social_accounts — the correct workspace is resolved automatically.

delete_post

Delete a scheduled or draft post. Published posts cannot be deleted.

Accounts & Analytics

list_social_accounts

List all connected social media accounts with IDs, platforms, and usernames.

list_analytics

Get post analytics — views, likes, comments, shares. Filter by platform.

sync_analytics

Trigger a fresh sync of analytics data from all connected platforms.

list_post_results

Check per-platform posting results — success/failure with error details.

Media Management

upload_media

Upload media and get back a reusable media_id. Pass a public URL or raw base64 data.

list_media

List uploaded media files with IDs, URLs, and types.

delete_media

Delete an uploaded media file (only if not used by any post).

Important

Platform Media Rules

The MCP Server strictly enforces platform media rules to ensure successful posting:

Platform Media Requirements
YouTube 1 video required
TikTok 1 video, or 1+ images required
Instagram 1–10 images/videos required
Pinterest 1 image or video required
Twitter / X Up to 4 images, or 1 video (optional)
Threads Up to 4 images, or 1 video (optional)
Bluesky Up to 4 images, or 1 video (optional)
LinkedIn Up to 20 images, 1 video, or 1 PDF (optional)
Facebook 1 video, or 1+ images (optional)
Good to Know

Multi-Workspace Support

PostSpread supports multiple workspaces. Simply pass the social account IDs from list_social_accounts — the correct workspace is resolved automatically.

No workspace_id is ever needed! The system automatically determines the correct workspace based on the social accounts you're posting to.

Package Info

NPM Package

The PostSpread MCP server is available as an npm package:

npm i postspread-mcp

View on npm →

Ready to get started?

Connect your AI agent to PostSpread and start managing social media from chat.