Last released Oct 20, 2026
AnyAISafe: audit, measure, price, apply and attest AI-safety controls for any software that calls a model or runs an agent; runtime guardrails included