Versions:
icuvisor is a local-first Model Context Protocol (MCP) server developed and published by Ricardo Niederberger Cabral, designed to connect training data from intervals.icu to AI assistants. Its primary purpose is to act as a bridge between an athlete's structured training information hosted on intervals.icu and conversational AI tools, enabling those tools to access and work with the user's training data through the standardized MCP interface. Because icuvisor is local-first, the server runs on the user's own machine, keeping the connection between the data source and the AI client under the user's local control rather than relying on a remote intermediary service. The software falls within the category of MCP servers and AI integration tooling, with relevance to sports, fitness, and endurance training workflows, since intervals.icu is a platform centered on training data. Typical use cases include allowing an AI assistant to reference an athlete's training history, plans, and performance information during a conversation, so that the assistant can respond with context drawn from the user's actual data. For example, a cyclist, runner, or triathlete using intervals.icu could query a compatible AI assistant about their training while the assistant draws on the connected data through icuvisor. The server is compatible with a range of MCP-capable clients, including Claude, ChatGPT, Pi, Cursor, and other MCP-compatible assistants, which gives users flexibility in choosing their preferred AI environment without changing the underlying connection to their training data. The current release of icuvisor is version 1.6.4, and the catalog records a total of 13 published versions, indicating an actively maintained project with a history of iterative releases. Users evaluating the software can consult the version history to track how the MCP server has evolved across its 13 releases leading up to version 1.6.4.
Tags: