Last released Jun 11, 2026
Intent-bound action authorization for AI agents — authorise before a credential is used, require human approval for sensitive actions, and prove every action traces back to the instruction that authorised it.
Supported by
