Run your social media from any AI assistant with one MCP server
Connect the AI assistant you already use to your Pallyy account and ask in plain language: see what's scheduled, draft and schedule posts, import media, and pin notes to the calendar. No code, on every plan, free included.
Works in the AI apps you already use
Claude connects in one click from its connector directory. ChatGPT and Cursor take the server URL and sign in with your Pallyy account. Gemini, Perplexity, Grok, Meta Muse, Le Chat, and any other MCP client connect with an API key.
Claude
Pick Pallyy from the connector directory and click Connect.
Signs in with your Pallyy accountClaude Code
One command adds the server, and /mcp signs you in.
Signs in with your Pallyy accountChatGPT
Add the Pallyy plugin from its ChatGPT listing and sign in.
Signs in with your Pallyy accountCursor
Add the server to your MCP config and click through the login.
Signs in with your Pallyy accountSet up the Pallyy MCP server in three steps
Copy the server URL, add it to your assistant, then sign in to Pallyy and choose what it's allowed to do. A couple of minutes, no technical setup.
Copy the server URL
Every app uses the same address, and Claude lists Pallyy in its connector directory. Nothing to install.
Add it to your assistant
Click Connect in Claude, or paste the URL into your app's connector or MCP settings.
Sign in and approve
Pallyy shows what the app can do. Pick the social sets it may reach and click Authorize.
Server URL
https://app.pallyy.com/mcp
How to connect each app
The short version for each app, using the server URL above, with the full steps in the docs. Claude, ChatGPT, and Cursor sign in with your Pallyy account; the rest take an API key.
Claude (web and desktop)
- 1Open the Pallyy listing in Claude's connector directory, or find Pallyy under Customize > Connectors.
- 2Click Connect.
- 3Sign in and approve on the Pallyy page. Pallyy is then available in every chat.
Claude Code
- 1Run the command below.
- 2In a Claude Code session, run /mcp, pick Pallyy, and authenticate.
- 3Approve on the Pallyy page that opens in your browser.
claude mcp add --transport http pallyy https://app.pallyy.com/mcp
ChatGPT
- 1Open the Pallyy plugin page in ChatGPT and add the plugin.
- 2Sign in and approve on the Pallyy page.
- 3Start a chat with the Pallyy plugin enabled. The docs cover adding it by URL if the listing is not available to you.
Cursor
- 1Open Customize in Cursor's sidebar and add a new MCP server, or edit ~/.cursor/mcp.json.
- 2Add the entry below and save.
- 3Click the server's login prompt to open the Pallyy page and approve.
{
"mcpServers": {
"pallyy": { "url": "https://app.pallyy.com/mcp" }
}
}Gemini CLI
- 1Create an API key under Settings > API keys with only the permissions Gemini needs.
- 2Run the command below with your key.
- 3Run /mcp in Gemini to confirm Pallyy is connected.
gemini mcp add --transport http \ --header "Authorization: Bearer pallyy_your_api_key" \ pallyy https://app.pallyy.com/mcp
Perplexity
- 1Create an API key under Settings > API keys.
- 2In Perplexity, open All settings > Connectors and click + Custom connector.
- 3Name it Pallyy, paste the server URL, and under Advanced set Authentication to API Key with your key.
- 4Add it, then enable Pallyy as a source in the prompt box.
Grok
- 1Create an API key under Settings > API keys.
- 2In Grok Bot, open Settings > Plugins and add a custom MCP server with the server URL.
- 3Add a header named Authorization with the value below, then attach Pallyy in a chat with @.
Authorization: Bearer pallyy_your_api_key
Meta Muse
- 1Create an API key under Settings > API keys.
- 2Ask Muse to create a custom connector for Pallyy's MCP server at the server URL, using a bearer token.
- 3Paste the key into Muse's credential prompt when it asks.
Le Chat
- 1Create an API key under Settings > API keys.
- 2In Le Chat, open Connectors > + Add Connector > Custom MCP Connector, name it Pallyy, and paste the server URL.
- 3Choose HTTP Bearer Token when asked and paste your key.
Any other MCP client
- 1Create an API key under Settings > API keys with only the permissions the assistant needs.
- 2Add the server URL to your client with the key as a bearer token.
- 3The assistant can do what the key allows. No sign-in step.
Authorization: Bearer pallyy_your_api_key
What you can ask your assistant to do
Once connected, the assistant works in your account the way you would. Some of the requests it handles:
See what's scheduled
Lists your social sets and posts for any date range, filtered by status or approval, with links into Pallyy.
- "What's going out for BeanHouse Café this week?"
- "Which posts are still waiting for client approval?"
Draft and schedule posts
One request creates a post for several networks, each with its own caption and media, as a draft or scheduled if you allowed publishing.
- "Draft an Instagram and LinkedIn post for Friday 9am with the latte photo."
- "Move Monday's post to Tuesday and shorten the caption."
Import images and videos
Give it a link to an image or video and it lands in the right media library, ready to use. In ChatGPT, an image generated in the chat can be imported directly.
- "Add this product photo to the library and use it in Thursday's post."
- "Show me the videos in the library from this month."
Pin notes to the calendar
Reminders, campaign dates, and recurring prompts become notes on the calendar.
- "Add a note on the 14th that the product launch is that day."
- "Remind the team every Monday for the next eight weeks to batch content."
The assistant can only do what you approved. Leave publishing off and it drafts while you publish. Disconnect it any time from Settings > Connected Apps.
Every tool the server exposes
17 tools backed by the REST API. Every connection can read; creating and changing needs the permissions you granted.
You decide what the assistant can touch
Permissions are set when you approve an app, visible in Settings, and revocable in one click.
Approve exactly what it can do
The consent page lists every permission requested. Viewing is always included; editing, publishing, uploading, and notes only when asked for.
Limit it to specific social sets
Choose which social sets the app may reach. Everything else stays invisible, handy for one connection per client.
Drafts unless you say otherwise
Publishing needs its own permission. Without it, the assistant drafts and you press publish.
Connected Apps in Settings
Every connected app shows its permissions, social sets, and last use. Disconnect revokes it immediately.
Post history names the app
Posts an assistant creates or changes show the app's name in post history.
Same rules as the app
Post limits, video and carousel availability, and approvals apply to every tool call, and the assistant is told why when something is refused.
Post to ten networks from one conversation
One post can target every connected network at once, with its own caption and media per network, exactly like the composer.
Questions, answered
What is an MCP server?
An MCP server is a service an AI assistant connects to through the Model Context Protocol. It hands the assistant a fixed set of tools with clear inputs. Pallyy's gives Claude, ChatGPT, Cursor, and other clients 17 tools covering social sets, posts, media, calendar notes, and Pinterest boards.
Which AI apps work with the Pallyy MCP server?
Claude on the web and desktop, Claude Code, ChatGPT, and Cursor sign in with your Pallyy account. Claude has Pallyy in its connector directory; ChatGPT takes the server URL as a plugin and Cursor as an MCP server. Gemini, Perplexity, Grok, Meta Muse, Le Chat, and any other MCP client connect with a Pallyy API key. Each app has its own setup page in the docs.
Do I need to be a developer to set it up?
No. Click Connect in Claude's connector directory, or paste one URL into another assistant's connector settings, then sign in to Pallyy and approve. Nothing to install and no code to write.
Which plans include the MCP server?
The Pallyy MCP server is included on every plan: Free, Pro, Agency, and Scale. Your plan limits, such as monthly posts and media allowances, apply to everything the assistant creates.
Can the assistant publish posts without me?
Only if you allow it. Publishing needs its own permission, shown on the consent page. Without it, the assistant creates drafts and you publish in Pallyy. It can never delete posts.
What can I ask the assistant to do?
Anything the 17 tools cover: list social sets, see posts for any date range, create or edit posts for one or several networks, import images and videos from a link, create or change calendar notes, and list Pinterest boards to pin to. In ChatGPT it can also import an image generated in the chat. It cannot read analytics or reply to comments and messages.
Which social networks can it post to?
The same ten as Pallyy: Instagram, Facebook, X, LinkedIn, Google Business, Pinterest, TikTok, YouTube, Threads, and Bluesky. One post can target several at once, each with its own caption.
Can it use an image ChatGPT generated?
Yes, in ChatGPT. Ask it to add the generated image to your library or use it in a post, and ChatGPT hands the file to Pallyy directly through the upload_media_file tool. Claude and Cursor have no way to pass a file from the chat, so with those you give the assistant a public link to the image instead.
How is this different from the Pallyy API?
They expose the same capabilities under the same permission rules. The API is for software you build, with API keys. The MCP server is for AI assistants, which sign in with your Pallyy account and get ready-made tools. Nothing to build or host.
How do I disconnect an app?
Open Settings > Connected Apps and click Disconnect. Access ends immediately, and you can reconnect later by approving again. Approving an app a second time replaces its permissions rather than adding another connection.
Does using the MCP server cost anything extra?
No. Posts the assistant creates count toward your monthly post limit like any other. The AI app may have its own subscription requirements for custom connectors.