MCP Overview

MCP Overview

The Model Context Protocol (MCP) is an open standard for exposing tools, resources, and context to AI models in a structured, interoperable way.

Why MCP?

MCP separates the tool definition layer from the model layer. Any MCP-compatible model can use any MCP server’s tools without custom integration code.

MCP in This Platform

You can connect any MCP-compatible server to your agents. The platform automatically discovers available tools and exposes them during agent runs.

What is MCP?

The Model Context Protocol (MCP) is an open standard that defines how AI models and agents connect to external tools and data sources. It separates the tool layer from the model layer, making any MCP-compatible tool available to any MCP-compatible model without custom integration code.

Think of MCP as USB for AI tools. Just as USB let any device connect to any computer with a single interface, MCP lets any tool connect to any AI agent with a single protocol.

How Queue uses MCP

Queue agents are native MCP clients. You can connect any MCP-compatible server to a Queue agent in your queue.yaml, and the agent will automatically discover and use its tools.

This means the Queue ecosystem of built-in tools and any third-party MCP servers — file systems, databases, GitHub, Slack, and more — are immediately available to your agents.

MCP server types

Local servers

Run as a process on the same machine as the Queue CLI or runtime. Great for development and for tools that need direct filesystem or network access.

Remote servers

Run as an HTTPS service. Production-safe and scalable. Connect to them from any Queue workspace by URL.

Connecting a server

mcp: servers: - name: my-tools url: https://tools.my-company.com/mcp auth: type: api_key key: ${MY_TOOLS_API_KEY}

Discovering tools

Queue automatically calls the MCP server’s tools/list endpoint at agent startup and injects the tool definitions into the agent’s context. No manual registration needed.