Learning Meta-Skills for Agent Harness Design in Test-Time AI4AI
Across Harness-Bench and NewtonBench, full-bank meta-skills improve macro-average performance by 8.95 percentage points over no-skill construction, and 12.02 points over direct delivery of the same bank to the Target.
Cheng Qian, Kun-Lun Zhu, Bei-Bin Li et al.
· 0 citations