Back to evaluations
Public evaluation
Governance Interrupt Test
Evaluates if LangGraph correctly pauses for human authorization when risk exceeds pre-set limits.
Evaluation type
task based
Challenge
Design Human-in-the-Loop Agentic Governance with LangChain and LangGraph
Difficulty
Intermediate
Rigor
Unspecified
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
1
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
Governance Interrupt Test
Submits high-value underwriting submission and checks if execution halts at human check node.
Input format
JSON submission object with insured_value_usd and property_type
Output format
JSON showing execution_state, human_review_required flag, and current_node