Model Context Protocol

The social media MCP server for your AI

postqueen's MCP server lets Claude, ChatGPT, Cursor and any MCP client draft and schedule posts across 30 channels from plain English.

Ask for drafts and they wait on your calendar. Ask for a time and it publishes then.

Start a 7-day trial for $0 Read the docs

Why use a social media MCP server?

Tool calling turns your scheduler into something an assistant can operate for you. Here is what that unlocks.

Just ask in plain English

No endpoints to memorize and no scripts to write. Describe the posts you want and when they should go out, and your assistant turns the words into scheduled content.

30 channels at once

One post tool reaches every channel on your account, so a single prompt can fan one idea out from X to LinkedIn to TikTok without the assistant learning a new tool per network.

Connect once with a link

Setup is a single URL with your API key inside. Paste it into Claude, Cursor or any MCP client and the postqueen tools are live in seconds.

Schedule ahead

Assistants can queue posts for exact dates and times, so a five minute conversation can fill next week's calendar or space a whole campaign out day by day.

Create images and videos

Ask for a visual and postqueen's AI image and video generation returns media ready to attach to the post, without leaving the conversation.

Your keys stay yours

Channel passwords never touch the assistant. Agents hold one postqueen API key that you can scope with OAuth and revoke at any moment.

Connect it in under a minute

One key, one endpoint, one config block. No SDK required.

Get your API key

Open Settings, then Developers in your postqueen dashboard and generate an API key. That one key is all an assistant needs, and revoking it instantly cuts off every agent that uses it.

Get your key

# In your postqueen dashboard
# open Settings, then Developers.
# Generate an API key and copy it:
 
YOUR_API_KEY
 
# Revoke it anytime to disconnect
# every agent at once.

Choose your endpoint

Put the key straight in the URL for clients that take a single link, or keep it in an Authorization header. OAuth capable clients connect through the /mcp-oauth endpoint with mcp:read and mcp:write scopes.

Endpoints

# Easiest: key in the URL
https://api.postqueen.ai/mcp/YOUR_API_KEY
 
# Or send it as a header
https://api.postqueen.ai/mcp
Authorization: Bearer YOUR_API_KEY
 
# OAuth clients: /mcp-oauth with the
# mcp:read and mcp:write scopes
https://api.postqueen.ai/mcp-oauth

Add it to your client

Paste one JSON block into Claude Desktop or Cursor, or run a single command in Claude Code. The postqueen tools show up in your very next conversation.

Claude Desktop / Cursor / Claude Code

{
  "mcpServers": {
    "postqueen": {
      "url": "https://api.postqueen.ai/mcp/YOUR_API_KEY"
    }
  }
}
 
# Claude Code, one command:
claude mcp add --transport http postqueen \
  https://api.postqueen.ai/mcp/YOUR_API_KEY

Connecting other clients

ChatGPT

1.Open Settings, then Connectors, and add a custom connector.
2.Paste the endpoint with your key in it, the same URL from step two.
3.Approve the tools when ChatGPT lists them, then ask for a post in any chat.

Codex

1.Add the server to your Codex MCP configuration, the same JSON shape as Claude Desktop.
2.Codex picks it up on the next session, no restart of anything else needed.

Gemini CLI

1.Add postqueen to the mcpServers block in your Gemini CLI settings file.
2.In the Gemini web app the same connection appears under Connected Apps, and the phone apps answer with those tools once it is on.

OpenClaw and Hermes

1.Both can use this endpoint, or skip MCP and run the postqueen CLI directly.
2.The CLI route needs no MCP client at all: install it, log in with the same API key, and the agent runs commands.

Then just talk to it

No commands to memorize. These are the kinds of requests the MCP tools handle end to end.

Fill a week

Plan a week of posts about our v2 launch and schedule one every morning at 9am on X, LinkedIn and Threads.

Queued 7 posts, one per morning at 9 AM, each tuned to its platform. They are in your calendar now if you want a final pass.

Cross-post a launch

Turn this changelog into a launch thread and cross-post it tonight at 7pm to X, Mastodon and Bluesky.

Drafted the thread and scheduled all three for 7pm, converted to each platform's native thread format. The drafts sit in your queue for review.

Make the media too

Write a post about our new pottery glaze, generate a square image for it, and put it on Instagram and Pinterest for Saturday morning.

Generated the image, wrote a caption for each platform and scheduled both for Saturday at 10 AM. Pinterest went to your Glazes board. Open either one to swap the image before it goes out.

The tools your assistant gets

Connect once and your client discovers these, plus one more the MCP runtime adds for talking to the agent directly. The same set powers the assistant inside postqueen, so nothing is held back from agents.

integrationList

Lists the channels connected to your account, optionally narrowed to one customer group. This is where an agent starts, because it needs the ids before it can post anywhere.

groupList

Lists your customer groups, so an agency agent can work on one client's channels without touching another's.

integrationSchema

Returns what a given channel actually accepts: character limits, media rules and the settings unique to it. An agent reads this before composing, rather than guessing and failing at publish time.

triggerTool

Fetches the platform-side values a post needs that nobody can type from memory, such as subreddit flair ids, YouTube playlists or Pinterest boards.

schedulePostTool

Creates the post. Content per channel, media, and either a publish time or a draft that waits on the calendar until you give it one.

generateImageTool

Generates an image to attach to a post, returned ready to use without leaving the conversation.

generateVideoOptions

Lists the video generation models available on your plan, with what each one needs.

videoFunctionTool

Resolves the extra inputs some video models require, a voice id being the common one, before generation runs.

generateVideoTool

Generates the video itself once the model and its inputs are settled.

uploadFromUrlTool

Pulls an image or video from a public URL straight into your media library, so an agent can reuse something that already exists on the web.

One server, 30 channels

The tool set stays the same whatever you connect. Every channel on your account is somewhere those tools can publish.

Explore every channel

F.A.Q.

Frequently asked questions

What is the Model Context Protocol?

MCP is an open standard that lets AI assistants discover and call external tools. postqueen ships a built-in MCP server, so any compatible assistant can drive your social scheduling the same way it uses its other tools.

What can an assistant actually do through it?

Draft and schedule posts, pick channels, set publish times, generate and attach images or video, and list or validate your connected channels. It covers the core scheduling workflow end to end, though a few dashboard-only screens stay in the app.

Which AI assistants work with it?

Any MCP client. Claude Desktop, Claude Code, Cursor, ChatGPT, Codex and Gemini CLI connect over HTTP with your API key, and open agents like OpenClaw and Hermes can use the same endpoint or the postqueen CLI. Gemini also takes the connector in its web app, under Connected Apps, and then answers with those tools in the Gemini phone apps.

Is it safe to let an agent touch my accounts?

Agents never see your channel passwords, and revoking the API key cuts access instantly. How much rein you give is set one instruction at a time: ask for drafts and they sit on the calendar with no time attached, ask for a scheduled post and it goes out at that time. Either way it is on the calendar first, so you can edit or delete it right up to its slot.

What does the MCP server cost?

Nothing extra. It is included with every postqueen plan and agent actions share your normal plan limits. Cloud plans start at $20 per month with a 7-day free trial, and unlimited posts are on every plan.

Can I self-host the MCP server?

Yes. postqueen is open source under AGPL-3.0 and the MCP server ships in the same free Docker image as the app, so self-hosted installs get the exact same endpoints on their own domain.

Ready to get started?

Queue this week's posts in minutes, or hand the whole thing to your favorite AI. postqueen keeps publishing either way.

Start a 7-day trial for $0 See pricing

Seven days free. Cancel in one click from settings.