capability-evolver
A self-evolution engine for AI agents. Analyzes runtime history to identify improvements and applies protocol-constrained evolution.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/autogame-17/evolverWhat This Skill Does
The Capability Evolver is a sophisticated meta-skill designed for OpenClaw agents to perform autonomous self-improvement. By leveraging the GEP (Genetic Evolution Protocol), the agent monitors its own execution logs, identifies bottlenecks or failure points, and iteratively refines its own codebase or memory structures. Unlike static scripts, the Evolver treats the agent's logic as a dynamic asset, allowing it to adapt to changing environments or unexpected edge cases without manual developer intervention. It acts as an internal audit and upgrade system, turning an agent from a rigid script into an evolving entity.
Installation
To integrate the Capability Evolver, run the following command in your terminal: clawhub install openclaw/skills/skills/autogame-17/evolver. After installation, you must register your node identity by running evomap.js. Follow the provided prompt to retrieve your node_id and claim code. Once registered, add the A2A_NODE_ID to your environment variables or your ~/.openclaw/openclaw.json config file to link your agent to the EvoMap network for protocol-compliant synchronization.
Use Cases
- Long-Running Autonomous Agents: Perfect for agents required to run indefinitely in production where human debugging is unavailable.
- Error Resilience: Agents that frequently encounter API timeouts or fluctuating data formats can use the Evolver to patch their own error-handling logic.
- Performance Optimization: Automatically refines memory management and logic paths to reduce execution time based on historical runtime performance.
Example Prompts
- "Analyze the last three days of execution logs and suggest patches for the repeated connection timeout errors."
- "Run the evolve cycle in review mode and present a summary of the proposed changes before applying them to the core logic."
- "Activate continuous evolution loop and monitor for memory leakage during high-load processing."
Tips & Limitations
- Review Mode is Mandatory for Production: Always use the
--reviewflag on critical systems to prevent the agent from introducing breaking changes into your production environment. - Environment Variables: Never hardcode credentials. Ensure
A2A_NODE_IDis managed through secure environment variables. - Scope: The Evolver excels at logic and structural repairs but should not be used to bypass system-level security protocols or sandbox constraints. Always ensure you have backups of your agent's source code before allowing autonomous modifications.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-autogame-17-evolver": {
"enabled": true,
"auto_update": true
}
}
}Tags
Flags: file-write, file-read, code-execution, network-access
Related Skills
designer-intelligence-station
Designer intelligence collection tool. Monitors 46 public sources (AI/hardware/mobile/design), dynamic quality-based filtering v2.1.8, generates structured daily/weekly reports. All data stored locally.
ai-workflow-red-team-lite
对 AI 自动化流程做轻量红队演练,聚焦误用路径、边界失败和数据泄露风险。;use for red-team, ai, workflow workflows;do not use for 输出可直接滥用的攻击脚本, 帮助破坏系统.
soul-weaver
AI Soul Weaver - 12 Curated Celebrity Templates + Custom Generation. Generate OpenClaw agent configurations through dialog or select from 12 world-famous tech & business leaders.
style-cloner
提供1-5篇参考文章 + 原始素材,AI 分析参考文章的风格特征, 将素材改写成同风格的成品文章,输出3个版本供选择,支持强度调节和迭代优化。
project-evaluator
描述一个项目想法,AI 从市场/技术/商业/风险四个维度系统评估, 输出评估报告、竞品速查、MVP建议,帮你决策「值不值得做」。