The Open Agent Leaderboard (huggingface.co)

Establishes a public leaderboard to benchmark and compare the performance of AI agents across standardized tasks and metrics. The leaderboard aims to promote transparency and drive innovation in AI agent development through open evaluation standards.