PROJECT_STRUCTURE.md

Project Structure

Overview

This project is a GUI simulator driven by large language models (LLMs) for testing and evaluating AI agents' behavior in desktop-like environments.

Repository Layout

mirror-gui/
├── README.md                    # Project introduction and usage
├── LICENSE                      # Open source license
├── requirements.txt             # Python dependencies
├── src/                         # Source code
│   ├── __init__.py
│   ├── core/                    # Core modules
│   │   ├── __init__.py
│   │   ├── simulator.py         # GUI simulator core
│   │   ├── agent.py             # AI agent implementation
│   │   └── models.py            # Data model definitions
│   ├── utils/                   # Utility modules
│   │   ├── __init__.py
│   │   ├── llm.py               # LLM client wrapper
│   │   └── config.py            # Configuration
│   ├── prompts/                 # Prompt templates
│   │   ├── __init__.py
│   │   └── prompts.py           # System prompts
│   ├── generation/              # Task generation
│   │   ├── __init__.py
│   │   ├── pipeline_generator.py # Blueprint generator
│   │   ├── scene.py             # Single-task synthesizer
│   │   ├── popup_scene.py       # Popup task synthesizer
│   │   ├── batch.py             # Bulk task generator
│   │   └── seed_apps.py         # Seed application list
│   ├── alignment/               # Thought alignment modules
│   │   ├── __init__.py
│   │   ├── safe_thought.py
│   │   └── safe_thought_popup.py
│   └── cli/                     # Command line interface
│       └── main.py              # Main entry point
├── data/                        # Data files
│   ├── task_blueprints.json
│   ├── tasks/
│   │   ├── generated_tasks/
│   │   └── generated_popup_tasks/
│   ├── logs/
│   │   ├── try_log/
│   │   └── try_popup_tasks_log/
│   └── datasets/
│       └── dataset/
└── scripts/                     # Utility scripts
     ├── help.sh
     └── check_pipeline.py