​
What Are MCPs?

MCPs (short for Model Context Protocol) are a powerful way to connect AI agents like Julius to external data sources and tools. Think of them as secure bridges that allow AI to access live information from databases, file systems, APIs, and other applications, enabling real-time integration with your existing tools and data rather than being limited to the AI’s training knowledge. Each MCP consists of:
  • Tools: Specific actions Julius can perform (e.g., search your files, query a database, send a Slack message)
  • Server: The application or service that makes these tools available to Julius
MCPs allow Julius to connect directly to your existing tools and services, expanding what it can do- like accessing your company’s data, automating tasks, or integrating with the software you already use.

​
How MCPs + Data Connector Work Together

Here’s how to set up and use MCP servers:
  1. Connect Your MCP Server
    • Go to your Julius workspace settings and add an MCP server via the Data Connector Settings Page
    • Authenticate using OAuth or API key, just like other data connectors
  2. Use the Available Tools
    • Once connected, Julius can access all the tools that MCP server provides
    • These tools work alongside your other data connectors - you can use them together in the same conversation
    • Julius will automatically know when to use MCP tools based on what you’re asking for
  3. Seamless Integration
    • MCP tools appear naturally in your Julius conversations
    • No special interface needed - just ask Julius to do something, and it will use the appropriate tools
    • Works with your existing data sources and workflows