Skip to content
Back to full roadmap
topiccore

Human-in-the-Loop Patterns

Approval gates, plan review, escalation — agent + human = safest production.

3 hours1 resources1 prereqs

Patterns:

  1. Plan review — agent generates plan, human approves/revises, then execute
  2. Action approval — approval before each destructive action (capability gating above)
  3. Escalation — low confidence or policy-violation suspicion → escalate to human
  4. Periodic checkpoint — long-running agent (hours/days) asks "shall I continue?" at intervals
  5. Post-hoc review — agent ran autonomously, results sampled + human review (10% sample)

LangGraph interrupt(): native HITL primitive. Agent resumes from saved state.

Prerequisites

Resources(1)