Human-in-the-Loop (HITL)

A human reviews and approves specific agent actions before they execute.

Why it matters

For high-stakes actions, you want a human checkpoint. HITL provides safety without removing automation.

In practice

In our Paperclip framework, 'approval_required' agents cannot execute without human sign-off.

Related terms

Human-on-the-Loop

A human supervises agents and intervenes only when thresholds are crossed.

Governance

Policies, controls, and processes that ensure responsible agent behavior in production.

Guardrails

Rules and constraints that prevent an agent from taking harmful or unauthorized actions.

Back to glossary