Currently, the Evaluation entity’s View in the Drupal AI Evaluations module displays raw data and metadata without additional processing. This makes it difficult for users to understand the evaluation’s outcome at a glance, as they have to manually sift through the raw information to discern key findings.
The View should instead provide a clear and concise summary of the evaluation, including the reasoning behind the thumbs up or thumbs down decision. By leveraging the recorded messages and the power of an LLM (Large Language Model), this data can be processed into a user-friendly format.
The View must display the following:
Utilize an LLM to:
Active
1.0
Code