Input
- Required Inputs:
session: Complete interaction log between user and agent showing all steps taken
Output
Result: Binary score (0 or 1)Reasoning: Detailed explanation
Interpretation
- 1: Task successfully accomplished intended goal
- 0: Task failed or couldn’t be completed