feat(eval): dataset-level evaluators + classification schemas, samples, e2e#1663

Draft

ajay-kesavan wants to merge 21 commits into

mainfrom

feat/classification-evaluator-types

Commits on Jun 19, 2026

Merge remote-tracking branch 'origin/main' into feat/eval-dataset-evaluators
ajay-kesavan
committed
Merge remote-tracking branch 'origin/main' into feat/classification-evaluator-types
ajay-kesavan
committed
refactor(eval): embed aggregator specs in per-datapoint evaluator configs

ajay-kesavan
and
claude
committed
Merge branch 'feat/eval-dataset-evaluators' into feat/classification-evaluator-types
ajay-kesavan
committed
feat(eval): wire sample classification evaluators to embedded aggregators

ajay-kesavan
and
claude
committed
refactor(eval): apply ponytail-review cleanup

ajay-kesavan
and
claude
committed
Merge branch 'feat/eval-dataset-evaluators' into feat/classification-evaluator-types
ajay-kesavan
committed
refactor(eval): apply ponytail-review cleanup (justification + demo)

ajay-kesavan
and
claude
committed
fix(eval): address adversarial-review feedback on dataset evaluators

ajay-kesavan
and
claude
committed
Merge branch 'feat/eval-dataset-evaluators' into feat/classification-evaluator-types
ajay-kesavan
committed
fix(eval): address adversarial-review feedback on classification samples

ajay-kesavan
and
claude
committed
fix(eval): address codex P1 + lint failures on dataset evaluators

ajay-kesavan
and
claude
committed
Merge branch 'feat/eval-dataset-evaluators' into feat/classification-evaluator-types
ajay-kesavan
committed
test(eval): drop fscore-duplicate test that conflicts with #1663 H2 validator

ajay-kesavan
and
claude
committed
fix(eval): publish aggregators in classification evaluator type schemas

ajay-kesavan
and
claude
committed

Commits on Jun 24, 2026

Merge remote-tracking branch 'origin/main' into feat/classification-evaluator-types
ajay-kesavan
committed

Commits on Jul 7, 2026

Merge remote-tracking branch 'origin/main' into feat/classification-evaluator-types
ajay-kesavan
committed