A highly optimized, visually stunning, and hands-free custom desktop environment built on Arch Linux and the Hyprland window manager. This project combines a modern Catppuccin-themed layout with a robust background voice automation engine, system maintenance daemons, and local AI (Ollama) synthesis.

The system coordinates speech processing, desktop events, and graphical indicators asynchronously:

graph TD
    User([🎙️ User Speech]) -->|Super + V / Wake Word| Mic[ALSA / PulseAudio Microphone]
    Mic -->|Audio Buffer| VA[Voice Assistant Engine]
    VA -->|STT: Google / Sphinx API| Parse{Command Router}
    
    Parse -->|Ollama Query| Ollama[Local Ollama: Llama3]
    Ollama -->|AI Text Output| TTS[Text-to-Speech: pyttsx3 / Espeak]
    
    Parse -->|System Match| Cmd[Shell Command Executor]
    Cmd -->|hyprctl dispatch| Hyprland[Hyprland Window Manager]
    Cmd -->|Waybar Toggle| Waybar[Waybar Panels]
    Cmd -->|System Script| Maintenance[Maintenance Agent]
    
    VA -->|Audio Feedback / Alert| Sounds[System Audio Cues]
    VA -->|Status / Log Output| Notification[notify-send Desktop Notifications]
    
    Daemon[Systemd User Daemon] -->|Manages Background| VA
    AGS[AGS Control Center] -->|Polls Status / Trigger| Daemon

• Instant Voice Orchestration: Perform window navigation, system adjustments, and menu actions hands-free.
• Local LLM Integration (Ollama): Route questions to local models via voice commands ("ask ai...") and receive spoken responses.
• Zero-Dependency Virtual Env Setup: Full Python package isolation to ensure clean system installations (PEP 668 compliant).
• AGS Control Center Panel: A GTK-based Control Center with a voice automation monitor, real-time daemon status indicator, and manual trigger button.
• Dynamic Waybar Toggle: Switch seamlessly between horizontal and vertical status bars.
• Automated Housekeeping Timers: Daily systemd timer that runs cache cleaning, package cache prunes (via paccache), and log prunes, displaying visual notifications of space freed.
• Premium Theme & Aesthetics: Cohesive dark aesthetic styled around Catppuccin Mocha colors.

To deploy this environment, you must be running Arch Linux with a working audio setup (PipeWire, PulseAudio, or ALSA).

Ensure git and sudo are installed before running:

sudo pacman -S --needed git

Follow these commands to deploy the environment on a fresh Arch Linux system:

# 1. Clone the repository
git clone https://github.com/wasifshaffaq/voice-driven-linux.git
cd voice-driven-linux

# 2. Make the installer executable
chmod +x install.sh

# 3. Run the master deployment script
./install.sh

1. Core Packages: Installs Hyprland, Waybar, Rofi (Wayland), Kitty, GNU Stow, system audio utilities, and nerd fonts.
2. Audio/STT Support: Installs portaudio (for python-pyaudio), espeak-ng (for offline TTS), pacman-contrib (for paccache), and alsa-utils.
3. AUR Helpers: Checks for and installs paru if not found.
4. AUR Packages: Deploys aylurs-gtk-shell (AGS), swww (wallpaper manager), libinput-gestures, and auto-cpufreq.
5. Config Symlinking: Uses stow to securely symlink configuration folders into your home directory (~/.config/).
6. Python Sandbox Environment: Creates a virtual environment in ~/.config/voice-assistant/venv/ and installs pip dependencies (SpeechRecognition, pyaudio, pyttsx3, requests).
7. Systemd Services: Loads and enables the background voice daemon (voice-assistant.service) and the daily maintenance timers (system-maintenance.timer).

The voice assistant can be triggered in three ways:

1. Hotkey (On-Demand Mode): Press SUPER + V (Windows key + V) on your keyboard. The system will play an audio cue and listen for a single command.
2. Wake Word (Daemon Mode): Say "computer" close to your microphone. The assistant will wake up and listen for your instruction.
3. Control Center Button: Open the Control Center with SUPER + C and click the Speak Command button.

All speech commands are mapped dynamically in commands.json.

Add your own voice triggers and system shell actions by editing the JSON mapping file:

nano ~/.config/voice-assistant/commands.json

Example command entry:

{
  "pattern": "^open browser|launch firefox$",
  "command": "hyprctl dispatch exec firefox",
  "response": "Launching Web Browser"
}

Verify command parsing and matching rules using the built-in test runner:

python test_voice_assistant.py

If the voice assistant background daemon is not responding, inspect the systemd user service logs:

systemctl --user status voice-assistant.service
journalctl --user -u voice-assistant.service -n 50 -f

• Ensure your microphone is active and not hardware-muted: alsamixer
• Check default audio inputs using wireplumber: wpctl status

├── dotfiles
│   └── .config
│       ├── ags                    # Aylurs GTK Shell configs (Control Center)
│       ├── hypr                   # Hyprland Window Manager configurations
│       ├── rofi                   # Application Menu styling
│       ├── systemd                # User Systemd Services & Timers
│       ├── voice-assistant        # Voice engine (Python, Commands configuration)
│       └── waybar                 # Status bar layouts (vertical and horizontal)
├── install.sh                     # System deployment and setup orchestrator
├── test_voice_assistant.py        # Python unit tests for the command logic
└── README.md                      # Documentation

This project is open-source and available under the MIT License.