Last released Mar 21, 2026
Universal feedback protocol and learning system for AI agents. Turn corrections into persistent, weighted rules.
Supported by