bookmarks
5 bookmarks by @supabase

Introducing Supabase Evals. Our benchmark for how well AI coding agents build with Supabase. We run agents like Claude Code, Codex, and Open Code against real tasks and score what they do.↗

·@supabase·Aug 1, 2026·tool·benchmark·claude-code·codex·supabase