fix(eval): patch fallback, teardown, run-batch.sh, baseline isolation - #23

Open

iamcxa wants to merge 9 commits into

mainfrom

fix/eval-patch-fallback

Commits on Apr 2, 2026

fix(eval): add fallback for patches that create new files

iamcxa
and
claude
committed
4a0cd02

Commits on Apr 3, 2026

fix(eval): teardown cleanup for files created by reverse patches

iamcxa
and
claude
committed
d0b8c16
fix(recce): reviewer must follow up on data_impact=potential before classifying

iamcxa
and
claude
committed
1c31d1a

Commits on Apr 4, 2026

Commits on Apr 5, 2026

feat(eval): add 3 new scenario types — grain cascade, false positive trap, date collapse

iamcxa
and
claude
committed
1c20456
fix(eval): drop stale prod schema before baseline runs in batch mode

iamcxa
and
claude
committed
4661fc6
fix(eval): address review — tighten patch fallback, trim scenario paths

iamcxa
and
claude
committed
4cc7621

Commits on May 22, 2026

refactor(recce-eval): v0.2.0 SKILL.md consolidation + script cleanup

iamcxa
and
claude
committed
551dc17