Agent Evaluator

Score and evaluate AI agent behavior across 5 dimensions: accuracy, efficiency, clarity, safety, and helpfulness.

Features

• 5-dimension scoring system
• Real-time behavior analysis
• Action history tracking
• Safety flagging
• MIT Licensed

Install

openclaw skills install skylv-agent-evaluator

Usage

The skill evaluates agent actions and provides structured feedback for continuous improvement.

License

MIT