Connect AI Assistants With MCP

The Model Context Protocol (MCP) is an open standard for connecting AI assistants and agents to tools and data. Anthropic introduced it in November 2024. Instead of building a custom integration for every assistant, a service can expose its data once through MCP and any compatible assistant can use it.

Yes. Apex® supports MCP, so AI assistants and agents can work with your Apex® content and data through a standard, secure interface instead of a custom integration. MCP access is enabled per account, so contact support to turn it on for your team.

MCP gives AI assistants a standard way to reach your Apex® content, so you do not have to build and maintain a separate integration for each tool. Because the interface is shared, you can change or add assistants more easily, and access can be managed in one consistent way.

Any AI assistant or agent that supports MCP connections can, in principle, connect to Apex® once MCP access is enabled for your account. Support for MCP varies between assistants and versions, so check your assistant's documentation for how to add an MCP connection.

Decide which content and data an assistant should reach, and name the person on your team responsible for that access. Then contact support to enable MCP access for your account, add the connection in your AI assistant and test it with a simple request. Review access regularly afterwards.

Contact Cyberdunes support at support@cyberdunes.com to enable MCP access for your Apex® account. Before you ask, decide which content and data assistants should be able to reach and who on your team will own that access, so the setup matches what you need.

Decide exactly which Apex® content and data an assistant should be able to reach, and leave out anything it does not need. Limiting access to what the task requires reduces risk and makes the connection easier to review. Write the decision down so the access owner can check it later.

Name one person on your team who is responsible for MCP access to Apex®. That person decides what assistants can reach, requests changes from support, and reviews access regularly. A single owner makes it clear who to ask when something needs to change or be removed.

After you add the connection in your AI assistant, test it with a simple request, such as asking the assistant to find or summarize a piece of your Apex® content. If the assistant returns the right information, the connection works. If not, check the connection settings and contact support.

Review MCP access to Apex® regularly and remove what is no longer needed, such as connections for tools your team has stopped using or data an assistant no longer requires. Keep one named owner for access, and contact support whenever the scope of access should change.