It lets users automate Arctic MUD sessions by coupling a full‑featured telnet client with a local LLM via LM Studio. The app parses ANSI output, builds token‑efficient prompts from a rolling screen buffer, and streams AI‑generated commands while persisting lessons in SQLite. Designed for MUD enthusiasts and AI hobbyists who want a self‑steering, real‑time controllable client. Compared to generic MUD clients, it adds built‑in AI orchestration, anti‑loop safeguards, and live steering, delivering a smarter, more autonomous experience.
View on GitHub →WilliamSmithEdward/mud.AI