Skip to main content
MCP (Model Context Protocol) servers extend the AI Assistant’s capabilities by connecting external tools and data sources. This allows the AI to interact with Jira, GitHub, Slack, and other services directly from the chat. MCP Servers page

​
Installing from Registry

The easiest way to add MCP servers is through the built-in registry.
1

Browse the Registry

Go to Settings > MCP Servers and click the Browse Registry tab. Search for servers by name (e.g., “atlassian”, “github”, “slack”).Searching for Atlassian Rovo in the MCP registry
2

Install a Server

Click Install on the server you want. Review the server details including the endpoint URL and transport type, then confirm.MCP server installation dialog
3

Authenticate (if required)

Some servers require OAuth authentication. After installing, go to the Installed tab and click the Test button — it will detect OAuth and show an Authorize button if needed.

​
Adding Custom Servers

For internal or custom MCP servers, use the Add Custom tab. Add Custom MCP server form Required fields:
  • Server Name — Unique identifier (e.g., my-internal-tool)
  • Server URL — The MCP endpoint (e.g., https://mcp.example.com/sse)
  • Transport Type — HTTP (Streamable HTTP) or SSE (Server-Sent Events)
Optional:
  • Headers — Add custom headers for authentication (API keys, Bearer tokens)
For servers that support OAuth, you can skip the headers and use the Test button on the Installed tab to trigger the OAuth flow.

​
Managing Servers

On the Installed tab, you can:
ActionDescription
Enable/DisableToggle the server on or off
TestVerify the server connection
RemoveUninstall the server

​
How MCP Extends the AI

When MCP servers are connected, the AI Assistant gains new tools. For example:
  • Atlassian Rovo — Search Jira issues, read Confluence pages
  • GitHub — Browse repositories, read issues and PRs
  • Slack — Search messages, post updates
The AI automatically discovers available tools from connected servers and uses them when relevant during conversations.

​
What’s Next?

AI Assistant

Use the AI with your connected tools

LLM Configuration

Configure the AI provider powering the assistant