feat: include Runtime operation context in Lab agent report

[hyeokjun32]

Summary

• Added optional --runtime-result input to Lab agent-runtime-report.
• Preserved Runtime health/error/event evidence in Agent Runtime Reliability Report context.
• Updated README/portfolio docs and tests for Runtime result operation evidence.

Validation

• poetry run python3 -m pytest tests/test_agent_runtime_report.py -q
• poetry run python3 -m pytest -q
• poetry run inferedgelab portfolio-demo-check

Notes

• Existing orchestration summary, guard_analysis, compare, and deployment decision contracts remain backward-compatible.
• Runtime evidence is optional; reports still work without --runtime-result.