Project Structure

EMBODIOS Project Structure

Overview

EMBODIOS is organized into several key components that work together to create an AI-powered operating system.

Directory Structure

embodios/
├── src/embodi/           # Main Python source code
│   ├── core/             # Core OS components
│   ├── builder/          # Container and image builders
│   ├── cli/              # Command-line interface
│   ├── runtime/          # Runtime environment
│   ├── models/           # Model management
│   └── installer/        # Bundle and ISO creation
├── kernel/               # Bare-metal kernel (C/Assembly)
├── docs/                 # Documentation
├── models/               # Model storage
├── scripts/              # Utility scripts
└── examples/             # Example Modelfiles

Key Components

1. Core OS (src/embodi/core/)

The heart of EMBODIOS:

• embodi_os.py - Main OS entry point
• kernel.py - Kernel operations
• runtime_kernel.py - Runtime management
• hal.py - Hardware Abstraction Layer
• inference.py - AI inference engine
• nl_processor.py - Natural language processing

2. CLI (src/embodi/cli/)

Docker-like interface for EMBODIOS:

• main.py - CLI entry point
• commands.py - Command implementations (build, run, bundle)

3. Builder (src/embodi/builder/)

Container and image management:

• builder.py - Image builder
• modelfile.py - Modelfile parser
• converter.py - Model format conversion

4. Runtime (src/embodi/runtime/)

Runtime environment management:

• runtime.py - Main runtime
• container.py - Container management
• image.py - Image handling

5. Installer (src/embodi/installer/)

Deployment tools:

• bundle/bundler.py - Bundle creator for bare-metal deployment
• iso/create_iso.py - ISO builder
• hardware/detect.py - Hardware detection

6. Kernel (kernel/)

Bare-metal kernel in C/Assembly:

• core/ - Kernel core (console, panic, interrupts)
• mm/ - Memory management (PMM, VMM, heap)
• ai/ - AI inference engine
• arch/x86_64/ - x86_64 architecture support
• include/ - Header files

Documentation

docs/
├── getting-started.md      # Quick start guide
├── architecture.md         # System architecture
├── modelfile-reference.md  # Modelfile syntax
├── hardware.md             # Hardware support
├── api.md                  # API reference
├── bare-metal-deployment.md # Deployment guide
└── performance-benchmarks.md # Performance data

Models Directory

models/
└── README.md            # Model storage info

Models are stored here when downloaded or converted.

Development Workflow

1. Python Development: Work in src/embodi/
2. Build Image: Use CLI embodi build
3. Test: Use CLI embodi run
4. Deploy: Create ISO with embodi bundle

Key Files

• setup.py - Package configuration
• requirements.txt - Python dependencies
• CONTRIBUTING.md - Contribution guidelines
• CHANGELOG.md - Version history
• README.md - Project overview

Architecture Notes

EMBODIOS follows a layered architecture:

1. Hardware Layer: Direct hardware access via memory-mapped I/O
2. HAL Layer: Unified interface for different hardware
3. Kernel Layer: Core OS functionality
4. AI Layer: Language model inference
5. Application Layer: Natural language interface