...An LLM agent searches connected indexers, selects a result according to quality and subtitle preferences, transfers it, and verifies what arrived. The library tracks season coverage, missing episodes, airing status, and scheduled gap-filling tasks. It supports 115, Quark, and GuangYaPan as separate workspaces without requiring local media downloads. Users can run an Electron desktop app or an always-on Docker deployment with browser-based configuration. A local agent API, live activity queue, notifications, and bring-your-own-model settings support both visual and automated operation.