Causal Safety Engine

Hi Product Hunt 👋 I built Causal Safety Engine after repeatedly seeing AI agents make decisions that were statistically confident but causally unsafe. Most systems validate actions using correlation-based signals. This works until you deploy agents in high-risk or autonomous settings, where spurious correlations, leakage, or unstable signals can silently cause failures. Causal Safety Engine is designed as a safety and governance layer, not a decision-maker: it checks whether an action is causally supported, stable, and identifiable — and prefers causal silence over false positives. I’m especially interested in feedback from people working on AI agents, safety, governance, or high-risk ML systems.