Personal Assistant

A voice-controlled AI assistant designed to run locally. This system leverages a local Large Language Model for intelligence and integrates with various external services and system tools to automate tasks and manage productivity.

Features

• Local Processing: Uses Ollama for local LLM inference and a local Text-to-Speech engine for low-latency voice interaction.
• Voice Interface: Real-time speech-to-text and text-to-speech capabilities.
• Tools & Integrations:
  • Google Workspace: Reads emails and manages calendar events.
  • System Control: Checks resource usage, manages files, and executes Python code in a sandbox.
  • Web Browser Control: Blocks distracting websites via a companion Chrome extension.
  • Memory: Maintains a persistent memory interface to recall user preferences and past interactions.
  • Self-Scheduling: Can schedule prompts to remind or check in with the user at future times.
  • Database: Syncs habit and tracking data with Supabase.

Setup

Prerequisites

• Python: Ensure Python 3.10+ is installed.
• Ollama: Install Ollama and pull the model specified in src/config.py (default: qwen3:4b-instruct-2507-q4_K_M).
• uv: This project uses uv for dependency management.

Installation

1. Clone the repository.
2. Install dependencies:

   uv sync

Configuration

1. Environment Variables: Create a .env file in the root directory with the necessary API keys (e.g., TAVILY_API_KEY, SUPABASE_URL, SUPABASE_KEY).
2. Google Credentials: Place your google_credentials.json file in the tools/ directory to enable Gmail and Calendar features.
3. Chrome Extension:
   • Open Chrome and navigate to chrome://extensions.
   • Enable "Developer mode".
   • Click "Load unpacked" and select the chrome-extension directory from this project.

Usage

To start the assistant:

uv run src/main.py

Creating an Alias

You can verify the assistant is running by saying the wake word ("Jarvis").

To run the assistant from any directory, you can create a shell alias. Replace <path-to-project> with the absolute path to this directory.

PowerShell Add this function to your PowerShell profile ($PROFILE):

function jarvis {
    uv run --directory "<path-to-project>" src/main.py
}

Bash / Zsh Add this alias to your .bashrc or .zshrc:

alias jarvis='uv run --directory "<path-to-project>" src/main.py'