A sensitive action is proposed
An agent or application reaches a moment where human input is required by policy.
Follow the reference trust journey and test its decisions using an interactive learning activity. This interface does not authenticate a person.
Click through a sample checkpoint. Each step explains the responsibilities of the person, the verifier, and the relying application.
Imagine an autonomous agent about to complete a consequential step. A human checkpoint helps make the handoff explicit.
An agent or application reaches a moment where human input is required by policy.
The interface explains the request and provides a clear, accessible way forward.
In a future implementation, a verifier would produce a time-bounded, context-bound result for server-side validation.
The relying application applies its own authorization rules and handles expiry, declines, and failures safely.
Explore the difference, the privacy implications, and why failure handling is part of a serious product.
Read the foundations