Agent Constitution Guard

Constitutional guardrails for AI agents — define immutable behavioral rules, permission boundaries, escalation policies, and safety interlocks that prevent a...