A web interface connecting local Ollama LLMs to Model Context Protocol (MCP) servers. Enables open-source models to use file operations, web search, and reasoning tools similar to commercial AI assistants - all running privately on your own hardware.