Repository navigation

feat(eve-bench): Terminal-Bench runner that separates invalid trials from scores - #3796

Closed

ruiconti wants to merge 10 commits into

mainfrom

rui/eve-terminal-bench

Closed

feat(eve-bench): Terminal-Bench runner that separates invalid trials from scores#3796
ruiconti wants to merge 10 commits into
mainfrom
rui/eve-terminal-bench

Commits on Sep 9, 2026
- feat(eve-bench): replace Harbor with a zero-dependency benchmark runner
  ruiconti
  committed
  d449b72
- feat(eve-bench): container hygiene, local task dirs, and stress scenarios
  ruiconti
  committed
  67d0a4c
Commits on Sep 25, 2026