Sync frontend step bugfix

Split architecture: Trainer=backend, Demo=dashboard

Add self-contained GRPO training notebook for HF Jupyter

Update README/Space card for hotseat multiplayer and 4 architectural upgrades

Add hotseat multiplayer Red Team controls and 4 architectural fixes

Gracefully handle invalid LLM actions with -0.2 penalty instead of 500 crash

Fix attack_chain KeyError and other validation errors

Initialize _step_reward_total in __init__ to fix AttributeError

Restore internal dashboard integration

Add root health check route to fix HF 404

Add FSP multi-agent architecture: Red Team LLM action space + alternating turns

Remove dashboard from backend

Fix business_impact grader: penalise unjustified isolations per-host

Add alternating self-play training scaffolding.

Replace SOAR playbooks with agentic micro-tools.

Add GRPO training pipeline + remove shield emoji

Fix radar chart size: switch scores panel to row layout

Add .gitignore, improve play_environment pending_followup tracking

Fix: stateful /demo endpoints, correct HF YAML, fix phantom data and 422 step errors

Fix phantom data bug: correct HF baseUrl, health endpoint for liveness probe, zero initial scores

Remove Hugging Face YAML metadata from README

Fix HuggingFace Space dashboard deployment (aiofiles, mount root)

Add HuggingFace Spaces YAML frontmatter

Initial commit of CyberSOC upgraded RLVR environment