Datasets:
Commit History
-
Mirror SkillsBench v1.1 as a benchmark task-tree dataset
Bingran You- 87 active task.md packages flattened to <task-id>/ at repository root - LFS rules for binary task assets - root eval.yaml (evaluation_framework: benchflow, task id skillsbench_v1_1) - dataset card aligned with the official benchflow-ai/skillsbench repod03762b