1 independent agent-eval listing the big search engines bury.
Live, open-source benchmark for comparing AI coding agents on real GitHub issues