Andrew Park 3fc0868106 fix(rescore_swe): preserve old score when new harness result is no_report ...

When upstream swebench's Modal cgroup-v2 patch can't be applied
(set_cpu_quota missing), the harness returns reason=no_report for the
rescored task. The previous code blindly overwrote rows[i].score with
new_score, destroying a previously-passing row. Now, if old score was
success=True and new is success=False+reason=no_report, keep the old.

Discovered 2026-05-26 after the rescore on advisors+skillorchestra
qwen36-opus47-swe-n100 cells inverted ~25 previously-passing rows to
no_report. Recovered via .predupe best-of dedup; next rescore will land
clean.

2026-05-26 01:47:01 -07:00

..

ablation

fix(rescore_swe): preserve old score when new harness result is no_report

2026-05-26 01:47:01 -07:00

install

Add cold-start CLI installer (#313)

2026-05-05 14:42:48 -07:00

mining

feat(mining): add Pearl model conversion workflow (#323)

2026-05-06 16:20:27 -07:00

bump-desktop-version.sh

refactor: merge desktop/ into frontend/, eliminate duplicate Tauri scaffolding (#246)

2026-04-14 09:48:12 -07:00

index_docs.py

Feature/twitter bot (#259)

2026-04-17 15:22:17 -07:00

oauth_all.py

fix(evals): clean up AI stack runtime harness (#320)

2026-05-05 19:28:50 -07:00

quickstart.sh

fix: address installation failures and SDK bugs (#100)

2026-03-24 16:29:26 -07:00

smoke_framework_comparison.sh

fix(evals): clean up AI stack runtime harness (#320)

2026-05-05 19:28:50 -07:00