Four composable agents that implement pre-execution governance gates, decision lineage recording, blast radius assessment, and kill path auditing for autonomous systems. Rule-based or LLM-backed. Ships with two pipelines: a governance pipeline (scope validation to execution to lineage) and a resilience pipeline (blast radius + kill path to a verdict). Part of a series on operational safety for AI systems.