xuanxuan

What This Skill Does

xuanxuan (旋旋) is a sophisticated skill-tracking and lifecycle management system designed for OpenClaw agents. It serves as an internal health monitoring engine that automatically logs, evaluates, and optimizes the performance of all your installed AI agent skills. By capturing real-time metrics—including execution success, response latency, and user satisfaction—xuanxuan transforms raw logs into actionable health scores. It acts as an automated 'SRE' for your AI environment, ensuring that your agentic workspace stays performant and efficient by identifying underperforming skills before they become bottlenecks.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/alsoforever/xuanxuan Once installed, you can integrate it into your own custom skills by importing log_skill_usage from the collect_usage module within your project scripts. Ensure your environment is configured for uv to handle dependency management automatically.

Use Cases

• Proactive Performance Tuning: Identify which skills are becoming slow or failing frequently, allowing for early refactoring.
• Agent Lifecycle Governance: Automatically generate monthly or weekly performance reports to audit which skills provide the most value.
• Automated Optimization Loops: Use the generated optimization proposals to prioritize maintenance tasks for your most critical agentic workflows.
• Deployment Monitoring: Verify if new skills or updates improve or degrade the overall health of your OpenClaw ecosystem.

Example Prompts

1. "xuanxuan, please analyze the health of all my currently installed skills and generate a summary report for the last 30 days."
2. "Run the optimizer for the 'code-review' skill and tell me why its performance score has dropped below 5."
3. "Show me the health status for all my skills and highlight any that are currently in the 'danger' or 'warning' categories."

Tips & Limitations

• Regular Analysis: While manual triggers work, it is highly recommended to configure the provided cron jobs to ensure your health scores remain up-to-date without manual intervention.
• Accuracy: The 'Satisfaction' metric relies on user-provided scores; ensure your integration hook includes this field whenever possible to improve the accuracy of the health-check algorithms.
• Storage Usage: Because xuanxuan stores logs in JSONL format, monitor your ~/.openclaw/data/gungun/ directory size if you track a high volume of agent calls daily.