wtf is uf-schedule-mcp?

This project is an MCP server, a small piece of software that gives an AI assistant like Claude structured access to a specific set of data or tools. In this case, the data is the University of Florida's public Schedule of Courses and published course syllabi. The README is clear that this is not an app you use directly. Instead, you just ask your AI assistant a plain English question, such as finding graduate statistics courses meeting on Tuesdays, and the assistant decides which functions to call behind the scenes. The server reads two public UF websites: the Schedule of Courses system, which lists courses and sections, and Simple Syllabus, which hosts published syllabi. Everything it reads is public information, requiring no login and no scraping of restricted pages. It offers four main tools. One searches courses using many filters like term, department, instructor, and credit hours. Another looks up a single course and all of its sections. A third lists valid filter option codes, such as which terms or departments exist. The fourth fetches a section's published syllabus content as structured data rather than a PDF file. The README is honest about real limitations. Meeting times and room numbers are hidden on the public schedule system unless you are logged in, though this information sometimes shows up instead through a section's public syllabus. Open seat counts are never available anywhere in the public data, so the server genuinely cannot report them. The syllabus tool returns structured content, not an official PDF, and only for syllabi marked publicly visible. Installing it is meant for people using AI assistants that support MCP, such as Claude Desktop, Claude Code, or Codex. Claude Desktop users can drag and drop a downloaded bundle file. Claude Code and Codex users run a single command that fetches and builds the server directly from GitHub. Other MCP compatible assistants can run it the same way over a standard local connection.