Skip to main content

Overview

Human-in-the-loop (HITL) nodes pause agent execution to collect human input. This enables:
  • Approvals - High-stakes decisions requiring human sign-off
  • Escalation - Complex cases beyond agent capabilities
  • Quality Control - Spot-checking agent outputs
  • Training Data - Collecting human feedback for improvement

Basic Configuration

Parameters

Approval Patterns

Simple Approval

Binary approve/reject for high-stakes decisions:

Multi-Level Approval

Chain multiple approval steps:

Escalation Patterns

Confidence-Based Escalation

Escalate when agent confidence is low:

Exception-Based Escalation

Escalate specific scenarios:

Timeout Handling

Configure what happens when humans don’t respond:

Escalation Chain

Context and Display

Provide rich context for human reviewers:

Response Handling

Structured Responses

Collect structured data from humans:

Response-Based Routing

Route based on human response:

Assignment Strategies

Role-Based Assignment

Round-Robin

Load-Based

Best Practices

Clear Prompts

Provide all context needed for the human to make a decision

Reasonable Timeouts

Set timeouts based on urgency and expected response time

Graceful Escalation

Always have a fallback when humans don’t respond

Feedback Loops

Use human responses to improve agent performance over time

Next Steps

Testing Agents

Test HITL workflows with mock responses

Node Types

Learn about other node types