OpenDoor AI

OpenDoor is a sophisticated, experimental coding assistant designed to bridge the gap between cloud-based AI (Gemini) and local LLMs (Ollama).

Important

The Future Agnet Vision: The ultimate goal of this project is to evolve from a "guided assistant" into a full-fledged Agnet.

• What is an Agnet? An AI Agnet is an autonomous agent capable of making decisions, solving complex architectural problems, and building entire projects independently.
• Current Status: The project is currently in the prototype stage and is INCOMPLETE. However, it contains the high-performance core and safety foundations required for its future evolution.

Installation

You can install OpenDoor directly from GitHub to use it as a global command on your system without downloading the source code manually:

pip install git+https://github.com/YASSER-27/opendoor.git

---

  ___                   ____                  
 / _ \ _ __   ___ _ __ |  _ \  ___   ___  _ __
| | | | '_ \ / _ \ '_ \| | | |/ _ \ / _ \| '__|
| |_| | |_) |  __/ | | | |_| | (_) | (_) | |   
 \___/| .__/ \___|_| |_|____/ \___/ \___/|_|   
      |_|                                       

Once installed, simply run opendoor in any terminal to start.

Detailed Features

1. Advanced Thinking & Transparency

OpenDoor provides a transparent "Thinking Mode" during response generation:

• Real-time Word Counter: Tracks every generated word as it happens.
• Microsecond Timer: Measures response latency to the millisecond for performance tracking.
• UI Throttling: The terminal UI is throttled at 15Hz to ensure zero CPU overhead on weak hardware during rapid text streaming.

2. Workspace Isolation & Safety

A strict isolation system is active during "Build Mode":

• Internal Blindness: Internal opendoor/ source files are hidden from the AI's context to prevent accidental self-modification.
• Sandboxed Extraction: Robust logic ensures files are extracted correctly into isolated project subdirectories.

3. C++ Performance Engine (The "Magic" Core)

OpenDoor is powered by a C++ performance core to ensure lightning-fast execution and minimal resource footprint:

• C++ Native Speed: The heavy-duty word counting and state tracking logic are architected with C++ principles (and a native C++ blueprint in engine/performance_core.cpp) to handle massive text streams without lagging.
• Incremental State Processing: Instead of re-processing large text buffers, OpenDoor uses O(1) incremental processing.
• System Call Caching: Terminal dimensions and TTY status are cached to minimize expensive Kernel-level requests.

4. Enterprise-Grade Portability

• Standalone Executable: OpenDoor can be compiled into a single (~56MB) .exe file that runs on any Windows machine without requiring Python.
• Bespoke Branding: Support for custom .ico assets embedded directly into the binary.

Technical Stack

• Languages: Python 3.14+ & C++ (Native performance core for critical path).
• AI Infrastructure:
  • Google GenAI SDK (Gemini Flash/Pro).
  • Ollama API (Local models like DeepSeek-R1, Llama 3).
• UI Architecture: ANSI-based Terminal User Interface (TUI) with optimized rendering.

Roadmap

• Transition to full Agnet autonomous architecture.
• Autonomous Command Execution: Ability to run tests and fix bugs without human intervention.
• Multi-modal generation support.

---

Built with passion by YASSER-27
AI Systems Architect & Visionary of the Agnet Evolution

Innovating towards the future of AI Agents.