fix: handle timezone-aware --until filtering in fetch-workflow-logs

[github-actions]

Summary

This fixes incorrect workflow run filtering when scripts/fetch-workflow-logs.py is called with a timezone-aware --until value (for example 2025-01-01T00:00:00+01:00).

The previous implementation compared created_at and boundary values as raw strings, which can misorder valid ISO-8601 timestamps with different timezone forms (Z vs +HH:MM).

Changes

• Parse run and boundary timestamps as datetimes and normalize comparisons in UTC.
• Keep date-only --until behavior inclusive by preserving end-of-day normalization.
• Treat empty timestamp strings as missing values to avoid boundary-check crashes.
• Add regression tests for:
  • timezone-aware --until cutoff behavior
  • empty created_at handling with an --until filter

Validation

• python -m pytest tests/test_fetch_workflow_logs.py -v
• python -m pytest tests/ -q

Closes #1236.

---

What is this? | From workflow: Trigger Bug Hunter

Give us feedback! React with 🚀 if perfect, 👍 if helpful, 👎 if not.

[v1v]

@copilot resolve the merge conflicts in this pull request

[Copilot]

@copilot resolve the merge conflicts in this pull request

Resolved. The conflict in scripts/fetch-workflow-logs.py was in list_workflow_runs — I merged both sides by keeping the datetime-based UTC comparisons from this branch while adopting the correct early-exit order (since-boundary check before conclusion filter) from main. All 6 tests pass.