From the course: Prometheus and Grafana: Visualizing Application Performance

Unlock this course with a free trial

Join today to access over 24,900 courses taught by industry experts.

Lab 2: Upgrading the assistant with MCP

Lab 2: Upgrading the assistant with MCP

From the course: Prometheus and Grafana: Visualizing Application Performance

Start my 1-month free trial Buy for my team

Lab 2: Upgrading the assistant with MCP

“

- [Instructor] Now, let's use the Grafana MCP server to provide context and access to our dashboard without writing specialized RAG code. MCP allows AI assistance to interact directly with Grafana, making the decisions smarter. Grafana has an official MCP server implementation that is open source and available on GitHub. It is still in early development, but we can use it locally to explore the power of MCP. There are several tools already like search_dashboards, get_dashboard_by_uid, query_prometheus and so on. When we configure the MCP server with Grafana credentials, it will be able to run commands that interact with our local Grafana instance on port 9090. The first thing we'll do is clone the repository. We'll then build the server command and remember the location. Next, we can configure our backend server to use MCP instead. We have changed the USE_MCP environment variable to true, and for the MCP server config, we have set the Grafana URL and the full path to the MCP server…

Contents