This Tempo MCP Server is a Go-based implementation of the Model Context Protocol (MCP) with native Grafana Tempo integration. It enables AI assistants and workflow tools to query and analyze distributed tracing data stored in Tempo using MCP tools defined by the server. The server follows the MCP specification to expose Tempo-specific tooling, notably the tempo_query tool, which accepts a required Tempo query string and optional connection details such as Tempo URL, time range, result limit, and authentication credentials. The server supports two modes of operation: a standard MCP interaction over stdin/stdout and an HTTP server that exposes a Server-Sent Events (SSE) endpoint for real-time streaming and an MCP endpoint for protocol messaging. The default HTTP port is 8080, configurable via the SSE_PORT environment variable. It can be built and run locally using Go or Docker, with a Dockerfile and optional Docker Compose for a complete test environment. The project includes testing support (run-client.sh) and is designed to integrate with Claude Desktop, Cursor, and n8n through configured MCP servers. Example configurations show how to expose tempo_query for Claude Desktop and other tooling to interact with Tempo data.