HF RL Explorer

internlm/WildClawBench-Harbor

WildClawBench-Harbor: Harbor dataset on Hugging Face with 60 tasks. All 60 WildClawBench tasks as Harbor task directories. Point harbor run at this folder and any Harbor agent (Claude Code, Codex CLI, OpenHands, Gemini CLI or your own) takes the benchmark with the official graders, in the official…

Tasks

internlm/WildClawBench-Harbor on the Hugging Face Hub