Optional
Raw text content to evaluate
Additional context
Expected/ideal output for comparison
The agent's input
The agent's output/response
Raw text content to evaluate