Evaluates and improves other agent skills

Summary

This skill assesses the quality of another agent's output against task requirements, providing structured JSON feedback. It identifies deficiencies, assigns a score (0.0 to 1.0), and specifies any missing instructions for improvement.