Grafana MCP Server

Grafana MCP Server

The Grafana MCP server allows you to interact with your Grafana instance through natural language, making it easier to manage dashboards, alerts, and administrative functions.

Prerequisites

• A running Grafana instance
• Grafana API key with appropriate permissions
• Go 1.20 or later (if building from source)

Installation

You can install the Grafana MCP server in several ways:

Using Docker

docker run -p 8080:8080 -e GRAFANA_URL=https://your-grafana-instance.com -e GRAFANA_API_KEY=your-api-key grafana/mcp-grafana

Building from source

1. Clone the repository:

git clone https://github.com/grafana/mcp-grafana.git
cd mcp-grafana

2. Build the binary:

make build

3. Run the server:

./dist/mcp-grafana --grafana-url=https://your-grafana-instance.com --grafana-api-key=your-api-key

Configuration

The Grafana MCP server requires the following environment variables or command-line flags:

• GRAFANA_URL or --grafana-url: The URL of your Grafana instance
• GRAFANA_API_KEY or --grafana-api-key: A valid Grafana API key with appropriate permissions

Additional optional configuration:

• PORT or --port: The port to run the MCP server on (default: 8080)
• LOG_LEVEL or --log-level: The logging level (default: info)

Adding to your AI assistant

To add the Grafana MCP server to your AI assistant, add the following configuration to your assistant's settings:

"mcpServers": {
  "grafana": {
    "url": "http://localhost:8080"
  }
}

If you're running the server with Docker or on a different host, replace localhost with the appropriate hostname or IP address.

Usage

Once configured, you can interact with your Grafana instance through natural language. Here are some examples:

• "Show me all dashboards"
• "Find dashboards related to CPU usage"
• "List all alert rules"
• "Show me firing alerts"
• "Create a new contact point for Slack"
• "List all teams in my Grafana instance"

The MCP server provides tools for managing dashboards, alerts, contact points, and administrative functions, making it easier to interact with your Grafana instance without navigating the UI.