In this session, we will explore mcp-nomad, a Go-based server implementing the Model Context Protocol (MCP) to interact with a Nomad cluster. This project enables dynamic access to jobs, nodes, logs, and system metrics via a unified and extensible interface. The aim is to demonstrate how we can bridge AI with real infrastructure, offering functionalities such as:

- Listing Nomad jobs and nodes

- Fetching allocation logs

- Manage Access Control List

- Collecting real-time system metrics

During the talk, we will discuss the architecture of the system, the challenges faced, and how MCP works, its underlying principles, and how it facilitates communication between systems.

A live demo will showcase mcp-nomad integrated into a real scenario. The session will conclude with reflections on potential future integrations and developments, including with other HashiCorp tools.