Eugene Peng

Back

Chess AI (Classical Search)

View on GitHub

Overview#

A chess engine built from scratch in Python with Claude Code as a learning project. It implements classical search techniques used by real chess engines — negamax with alpha-beta pruning, quiescence search, MVV-LVA move ordering, and tapered PeSTO evaluation — and is playable via a web interface or the UCI protocol against other engines.

Playing strength: ~1580–1640 ELO (measured vs Stockfish with UCI_LimitStrength)