a minimal chess engine built with Go that implements negamax search with alpha-beta pruning, material evaluation, and full move validation for checkmate, stalemate, and rule compliance under constrained rules. Designed to fit within 1KB of code, it emphasizes readability, correctness, and educational value for learning AI-driven game engines. Supports Elo rating via CI games against Stockfish at fixed depth, with move ordering and capture-first strategy for performance. Lightweight, secure and deployable via Docker and Kubernetes for AI-native development workflows. ai-engineering, go, ai-agents GitHub