Add generic Harbor benchmark runner#755

Merged

neubig merged 8 commits into

mainfrom

feat/generic-harbor-eval

Jun 23, 2026

Commits on Jun 22, 2026

Add generic Harbor benchmark runner

Graham Neubig
and
openhands-agent
committed
Fix LaminarService call in eval_infer to use correct method and make telemetry non-fatal

Graham Neubig
and
openhands-agent
committed
fix: resolve ruff format/lint and pyright issues in harbor eval (#755)

Graham Neubig
and
openhands-agent
committed
fix: match CI ruff formatting and remove unused variable (#755)

Graham Neubig
and
openhands-agent
committed
fix: address review comments on harbor run_infer (#755)

Graham Neubig
and
openhands-agent
committed
fix: pyright type error in test helper (#755)

Graham Neubig
and
openhands-agent
committed
fix: address 2nd round review — adapter reproducibility + secret masking (#755)

Graham Neubig
and
openhands-agent
committed
fix: extend secret masking to --ak kwargs (#755)

Graham Neubig
and
openhands-agent
committed