Gate max queue depth in Runtime Intelligence summary

[hyeokjun32]

Summary

• update Runtime Intelligence fixtures so preserved Orchestrator operation context includes max_total_queue_depth
• gate Lab Markdown/HTML Runtime Intelligence Risk Summary output on max_total_queue_depth in queue/deadline/fallback markers
• document that Lab prefers max_total_queue_depth when present while keeping the row as reviewer navigation context

Tests

• python -m pytest -q tests/test_report_generators.py tests/test_compare_service.py tests/test_runtime_intelligence_evidence_chain_smoke.py tests/test_runtime_intelligence_ci_template.py -p no:cacheprovider
• poetry run python3 -m pytest -q -p no:cacheprovider
• poetry run inferedgelab portfolio-demo-check
• INFEREDGE_REPOS_DIR=/Users/GwonHyeokJun/Documents/GitHub INFEREDGE_AGENT_RUNTIME_EDGEENV_SMOKE_OUT=/private/tmp/inferedge_agent_runtime_lab_max_queue_marker_20260601 INFEREDGE_RUNTIME_INTELLIGENCE_SMOKE_OUT=/private/tmp/inferedge_runtime_intelligence_lab_max_queue_marker_20260601 INFEREDGE_REMOTE_FALLBACK_REGISTRY_SMOKE_OUT=/private/tmp/inferedge_remote_fallback_lab_max_queue_marker_20260601 bash scripts/smoke_all.sh