This MCP server is a Model Context Protocol (MCP) server that provides voice status updates using OpenAI's text-to-speech API. It enables language models to communicate with users through brief audio messages, reporting on the progress of a task or confirming that a command has been executed. It is especially useful when working with Cursor or Claude code: you can assign a task to an agent, step away, and still receive audible progress updates and a notification when the task completes. The server is designed to be batteries included: MCP clients typically prepend the summarize tool to the system message, and this server uses that mechanism to prompt the model to report progress via speech. The available tooling includes the summarize(text: str) function, which converts the provided text into speech using OpenAI's TTS API and plays it back to the user.