Fix skipped BCal aggregation

[martinsrui-msft]

Summary

• always run final BCal aggregation, even when an entry job fails
• report expected, produced, missing, unexpected, and duplicate entries, while preserving diagnostic artifacts for incomplete runs
• block scoring, Kusto/Braintrust, and leaderboard publication when results are incomplete, and fail the workflow visibly
• isolate rerun artifacts by attempt and pin/cache BcContainerHelper once

This is the top layer above #943, split from #925. It intentionally adds no BCal execution retries.

Validation

• focused pytest suite: 116 passed
• Ruff check and format checks passed
• full bcbench Ty check passed
• relevant file-scoped pre-commit hooks passed

[github-code-quality]

Code Coverage Overview

Languages: Python

Python / code-coverage/pytest

The overall line coverage in commit d05f8ca in the martinsrui-bcal-resi... branch remains at 86%, unchanged from commit 973eb05 in the martinsrui-lm-checkl... branch.

Show a line coverage summary of the most impacted files.

File	martinsrui-lm-checkl... 973eb05	martinsrui-bcal-resi... d05f8ca	+/-
src/bcbench/commands/result.py	93%	89%	-4%
src/bcbench/results/display.py	93%	92%	-1%
src/bcbench/results/summary.py	96%	96%	0%
src/bcbench/res...lts/__init__.py	100%	100%	0%
src/bcbench/res...ceval_export.py	100%	100%	0%
src/bcbench/com...ands/dataset.py	25%	26%	+1%
src/bcbench/res...completeness.py	0%	96%	+96%

---

_{Updated October 07, 2026 08:20 UTC}