Guardian AI Package

This is the main package for the Guardian AI system, a robust, self-improving, self-healing, and self-preserving agent framework designed with security and simplicity in mind.

Package Structure

guardian_ai/
├── analysis/       # Code analysis and diagnostics
├── config/         # Configuration files
├── core/           # Core agent functionality
├── data/           # Data storage
│   └── backups/    # Backup files
├── docs/           # Documentation
├── interface/      # User and API interfaces
├── memory/         # Persistent memory management
├── monitoring/     # Monitoring functionality
├── recovery/       # System recovery
├── resilience/     # Fault tolerance mechanisms
├── security/       # Secure execution environment
│   └── validation/ # Input validation
└── tools/          # Utility tools for development

Documentation

For detailed documentation, see the docs/ directory:

• API Documentation
• Architecture Overview
• Security Guidelines
• File Naming Convention

Main Components

• Core Components

  • main.py: Main entry point for the agent framework
  • run_guardian.py: Command-line interface script
  • core/perpetual_llm.py: Primary driver of the agent framework
  • core/rsi_module.py: Handles recursive self-improvement
  • core/system_task_manager.py: Task scheduling and management

• Interface Components

  • interface/hitl_interface.py: Human oversight interface
  • interface/ollama_agent.py: Integration with Ollama LLM

• Security Components

  • security/sandbox_executor.py: Secure code execution with Docker isolation
  • security/validation: Input and output validation

• Memory Components

  • memory/memory_manager.py: Persistent state management with file monitoring and versioning

• Resilience Components

  • resilience/circuit_breaker.py: Circuit breaker implementation for resilience
  • resilience/retry_handler.py: Retry mechanisms for transient failures

• Monitoring Components

  • monitoring/dashboard.py: Web-based monitoring dashboard
  • monitoring/monitor.py: Performance and behavior monitoring
  • monitoring/health_checker.py: System health verification
  • monitoring/metrics_collector.py: Metrics collection with Prometheus

• Recovery Components

  • recovery/system_recovery.py: System backup and recovery

• Analysis Components

  • analysis/consolidated_code_analysis.py: Code analysis tools
  • analysis/self_diagnostic.py: Self-diagnostic capabilities

Usage

For usage instructions, see the main README.md file in the repository root.