Open leaderboard for browser agents
3 files for Claude Code, Codex and OpenCode: benchmark-discovery, leaderboard AGENTS.md, model-classification-research — 630 tokens loaded in every session.
AGENTS.md A 630 tok .claude/skills/benchmark-discovery/SKILL.md A 235 tok .claude/skills/model-classification-research/SKILL.md A 46 tok These files are steel-dev/leaderboard's own configuration — they tell Claude Code, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.