This MCP server enables AI assistants to query Azure Data Explorer (ADX/Kusto) databases and Eventhouse clusters via standardized MCP interfaces. It supports executing arbitrary KQL queries against a configured ADX database and returning results in JSON format for easy consumption. Discovery capabilities include listing tables, viewing table schemas and column types, sampling table data with configurable sizes, and retrieving detailed table statistics such as row counts and storage metrics. Authentication is provided via DefaultAzureCredential (Azure CLI, Managed Identity, etc.) and native Workload Identity for AKS environments, with a token file path option; if workload identity variables are not present, the server falls back to DefaultAzureCredential. Deployment options include multiple transports (stdio by default, HTTP, and Server-Sent Events), Docker container deployment, and a development Dev Container experience. The set of tools exposed by the MCP server is configurable, allowing you to tailor the MCP client surface to only the functionality you need for your use case.