NuHF Claw: A Risk Constrained Cognitive Agent Framework for Human Centered Procedure Support in Digital Nuclear Control Rooms

Build a risk-constrained agent framework specifically for safety-critical industrial environments. Focus on 'circuit-breaker' logic that prevents erratic LLM behavior.

Suggested repo: GuardClaw

"Safety-first agent framework for when 'hallucination' isn't an option."

Estimated effort: 100h