Find where your LLM agent's answers fall short.
Produces MLflow evaluation results and recommendations for improving an existing LLM agent.
When to use it
Use when evaluating or improving an existing LLM agent's answer quality, tool selection, costs, or incomplete responses.
Give it an existing agent project; it asks about the agent's purpose and failures, then produces evaluation results and an improvement report.
What you provide
Uses Python to run the evaluation and validation scripts.
Requires uv for MLflow CLI commands and Python script execution.
This skill
MLflow
Installs MLflow for evaluation
Records evaluation data in MLflow
https://mlflow.org/docs/latest/llms.txt
Consults current MLflow documentation