feat: surface sustained runtime reliability evidence

[hyeokjun32]

Summary

• Added sustained queue-depth and policy-decision context to Agent Runtime Reliability reports.
• Preserved AIGuard sustained_overload_risk evidence in Lab report output and deployment decision context.
• Updated committed agent-runtime fixtures and portfolio docs for the sustained high-load scenario.

Validation

• git diff --check
• poetry run python3 -m pytest tests/test_agent_runtime_report.py -q
• poetry run inferedgelab agent-runtime-report --orchestration-summary examples/agent_runtime/agent_3_orchestration_summary.json --guard-analysis examples/agent_runtime/aiguard_runtime_guard_analysis.json --format markdown
• poetry run inferedgelab portfolio-demo-check
• poetry run python3 -m pytest -q

Notes

• Lab remains the final deployment decision owner.
• AIGuard and Orchestrator remain optional evidence providers.
• No production cloud orchestration, DB/queue/auth/billing, or AI OS behavior was added.