Input
- Required Inputs:
session: Complete interaction log between user and agent showing all steps takenexpected_steps: List of required steps (order flexible)
Output
Result: Binary score (0 or 1)Reasoning: Detailed explanation of step completion
Interpretation
- 1: All steps completed (order flexible)
- 0: Missing steps or dependency violations