AgentEvolver Self-Improvement Plugin

Implements AgentEvolver's self-evolving mechanisms for Agent Zero AI agents, enabling autonomous self-improvement through three key mechanisms:

🧠 Three Self-Evolving Mechanisms

1. Self-Questioning (Automatic Task Generation)

Agents can generate their own improvement tasks based on performance gaps and learning needs.

2. Self-Navigating (Experience-guided Exploration)

Agents learn from past experiences and apply that knowledge to navigate new tasks more effectively.

3. Self-Attributing (Credit Assignment)

Agents analyze which actions contribute to success/failure for fine-grained policy optimization.

🚀 Features

• 8 Self-Questioning Tools: Complete toolset for task generation, experience logging, and improvement analysis
• Web UI Settings: Configurable parameters for each self-evolving mechanism
• Experience Database: Persistent storage of learning experiences in JSON format
• Performance Statistics: Track success rates, tasks completed, and learning progress
• Improvement Suggestions: AI-powered recommendations for skill development

📁 Plugin Structure

agentevolver_self_improvement/
├── plugin.yaml              # Plugin manifest
├── README.md                # This file
├── LICENSE                  # MIT License
├── hooks.py                 # Plugin lifecycle hooks
├── test_plugin.py           # Test suite
├── webui/
│   └── config.html          # Settings UI
├── helpers/
│   ├── __init__.py
│   └── self_improvement.py  # Core engine (274 lines)
└── tools/
    ├── __init__.py
    └── self_questioning_tool.py  # Tool interface (157 lines)

🛠️ Installation

Method 1: Plugin Hub (Recommended)

1. Open Agent Zero UI
2. Go to Plugins → Plugin Hub
3. Search for "AgentEvolver"
4. Click "Install"

Method 2: Manual Installation

# Clone the repository to your plugins directory
cd /a0/usr/plugins/
git clone https://github.com/your-username/agentevolver_self_improvement.git

⚙️ Configuration

After installation, configure the plugin via Settings → Agent → AgentEvolver Self-Improvement:

• Self-Questioning: Enable/disable automatic task generation
• Self-Navigating: Set experience pool size and learning parameters
• Self-Attributing: Configure credit assignment depth
• Auto-scheduling: Set up automatic self-improvement sessions

🧪 Usage Examples

Generate a self-improvement task:

{
  "tool_name": "self_questioning_generate_task",
  "tool_args": {
    "category": "coding",
    "difficulty": "medium"
  }
}

Record a learning experience:

{
  "tool_name": "self_questioning_add_experience",
  "tool_args": {
    "task_type": "research",
    "task_description": "Study AI agent frameworks",
    "actions": ["Search", "Read", "Summarize"],
    "outcome": "success",
    "lessons_learned": ["Found 5 frameworks", "Identified key features"]
  }
}

Get improvement suggestions:

{
  "tool_name": "self_questioning_get_improvement_suggestions",
  "tool_args": {}
}

📊 Performance Tracking

The plugin maintains detailed statistics:

• Success rate across task categories
• Total experiences and tasks completed
• Learning progress over time
• Pattern analysis for improvement areas

🔧 Development

Running Tests

cd /a0/usr/plugins/agentevolver_self_improvement
python test_plugin.py

Extending the Plugin

1. Add new task categories in helpers/self_improvement.py
2. Create new tools in tools/ directory
3. Extend the Web UI in webui/

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Submit a pull request with tests

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgements

Based on the AgentEvolver research paper:

"AgentEvolver: Towards Efficient Self-Evolving Agent System" Yunpeng Zhai et al., 2025

This plugin adapts the three self-evolving mechanisms for the Agent Zero framework.