Last released Oct 8, 2026
A small governance core for AI agents that take real actions: allowlisted capabilities, validated model output, human confirmation tokens and a hash-chained audit log.