Most creator tools are built for someone sitting at a desk with twenty minutes to click through menus. JubJub assumes you have neither.
This page covers three things: what the connection actually is, what you need to set it up, and what your AI can do once it is running.
What the MCP server actually is.
MCP stands for Model Context Protocol. It is an open standard that lets AI assistants connect to external tools and take real actions, not just look things up.
JubJub ships a production MCP server with 100+ tools. Publishing, analytics, splits, ownership registration, collaborator management, agent payments — all of it is accessible to your AI through a single connection.
The key word is actions. When your AI is connected to JubJub, it does not summarise what you could do. It does it. See how JubJub works as a conversational interface.
It does not summarise what you could do. It does it.
What you need to get set up.
Three things, and that is the whole list.
A JubJub account
The account that holds your catalogue, your splits, and your payment rails. Free to set up.
An AI assistant that supports MCP
Claude is the primary integration. Any assistant that speaks the Model Context Protocol can connect.
The JubJub MCP server URL
Added to your MCP client settings. The URL lives in your JubJub developer settings.
No code required for standard setup. The developer docs walk through the full process end to end.
What your AI can do once connected.
Once the connection is live, your AI can:
- Publish content across platforms in a single instruction.
- Pull analytics and summarise performance in plain language.
- Manage collaborator splits and workspace membership.
- Confirm ownership registration at publish.
- Handle agent payment flows on your behalf.
That is the short list. The full tool catalogue is in the developer docs. For a breakdown of what the AI actually does day to day, the companion page covers it in detail: what an AI agent can do for a creator.
The Anthropic partnership.
JubJub is part of the Claude Partner Program. The MCP integration is maintained against current specifications and ships updates as the protocol evolves.
Why MCP beats a custom integration.
A custom API integration means code, a developer, and a maintenance burden every time something changes on either side. MCP is a standard.
Your AI connects once. Any MCP-compatible AI assistant can use the same connection. If you change your AI setup later, the JubJub integration does not need rebuilding.
A custom integration is a maintenance burden. A standard is something you wire up once and forget about.





