zhua-evolver

What This Skill Does

The Zhua Evolver (zhua-evolver) serves as the autonomous cognitive engine for the OpenClaw AI agent. It is designed to bridge the gap between current operational capabilities and desired intelligence targets. By leveraging a structured five-step lifecycle—Gap Analysis, Skill Discovery, Automated Deployment, Evolution Logging, and Minion Orchestration—this skill allows the agent to self-optimize continuously. It goes beyond simple task execution by treating intelligence as an iterative process, utilizing a specialized team of sub-agents (the 'Claw Minions') to handle distributed responsibilities from security monitoring to technical documentation updates. This ensures the agent does not merely accumulate features but grows in a coherent, tracked, and secure manner.

Installation

To integrate the Zhua Evolver into your environment, use the OpenClaw standard package manager:

clawhub install openclaw/skills/skills/beipian261/zhua-evolver

Ensure that your environment supports Python 3 execution as the backend scripts for analysis and installation require it to interact with the system files and remote skill repositories.

Use Cases

• Continuous Learning: Use this when you want your agent to proactively discover new capabilities without manual intervention.
• Intelligence Upgrading: When the agent needs to move from a basic automation tier to an advanced problem-solving tier.
• Adaptive Refinement: When specific task success rates drop, the system triggers gap analysis to identify missing logic or modules.
• Scale Operations: When managing complex workflows, the orchestrator coordinates specialized sub-agents to maximize efficiency across multiple technical fronts.

Example Prompts

1. "爪爪，根据当前的反馈表现，运行自我进化循环，优先补强数据分析相关的技能缺口。"
2. "我们要达到进阶智能水平，请指挥小弟们搜索并安装最新的网络优化插件，并更新进化日志。"
3. "爪爪，启动全系统评估，通过差距分析识别并解决目前任务覆盖率低的问题。"

Tips & Limitations

• Safety First: The '守爪' (Guard Claw) minion is crucial. Always verify the log files after an automated installation to ensure no unauthorized configurations were applied.
• Resource Management: Constant evolution cycles can consume significant system resources. It is recommended to schedule large-scale evolution tasks during low-activity windows.
• Governance: Evolution is inherently iterative. Monitor the 'SOUL.md' file managed by the '魂爪' (Soul Claw) to understand how the agent’s identity and priority matrix are evolving over time. Avoid running multiple concurrent evolution cycles to prevent file system conflicts.