Surface AIGuard max queue traceability in Lab reports

[hyeokjun32]

Summary

• add a Runtime Intelligence Risk Summary row that ties the visible Lab max_total_queue_depth marker to AIGuard raw context
• gate the new row in Lab artifact and optional CI artifact checks
• update Runtime Intelligence docs and tests for the reviewer-facing report marker

Validation

• poetry run 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_artifact_gate.py tests/test_runtime_intelligence_ci_template.py -p no:cacheprovider
• poetry run python -m pytest -q -p no:cacheprovider
• poetry run inferedgelab portfolio-demo-check
• bash scripts/smoke_runtime_intelligence_chain.sh --output-dir /private/tmp/inferedgelab_max_queue_raw_context_report_row_20260601
• INFEREDGE_REPOS_DIR=/Users/GwonHyeokJun/Documents/GitHub INFEREDGE_AGENT_RUNTIME_EDGEENV_SMOKE_OUT=/private/tmp/inferedge_agent_runtime_lab_max_queue_raw_context_row_20260601 INFEREDGE_RUNTIME_INTELLIGENCE_SMOKE_OUT=/private/tmp/inferedge_runtime_intelligence_lab_max_queue_raw_context_row_20260601 INFEREDGE_REMOTE_FALLBACK_REGISTRY_SMOKE_OUT=/private/tmp/inferedge_remote_fallback_lab_max_queue_raw_context_row_20260601 bash scripts/smoke_all.sh