Zeta is a privacy-first, autonomous AI agent that runs 100% locally on your machine. It combines a secure terminal interface with a modern, "One UI 8" inspired Web Dashboard.

Unlike cloud agents, Zeta stores all memories, logs, and data on your device. It connects to Ollama for intelligence (Llama 3, DeepSeek, etc.) and uses ChromaDB for long-term semantic memory.

• Natural Language Understanding: Speak naturally (e.g., "Search for SpaceX news" or "Check NVDA stock").
• Long-Term Memory: Remembers facts about you and past conversations across sessions using Vector Search (ChromaDB) and SQL.
• Local Inference: Powered by Ollama. Zero data leaves your machine unless you ask for a Web Search.

1. Secure Terminal CLI: Fast, keyboard-centric interface for developers.
2. Modern Web Dashboard:
   • System Stats: Real-time CPU/RAM monitoring.
   • Memory Visualization: See how many facts Zeta has stored.
   • Chat Interface: Beautiful, fluid chat experience.

• Web Search: Uses DuckDuckGo to find real-time info.
• Finance: Live stock market data.
• System Control: Can execute terminal commands (with safety confirmations).
• Note Taking: Saves thoughts and ideas to local files.

• Python 3.10+
• Node.js & npm (for Web Dashboard)
• Ollama (Running locally with a model like llama3 or deepseek-r1)

1. Clone the Repository

   git clone https://github.com/pronzzz/zeta.git
   cd zeta

2. Run the Installer This script sets up the Python virtual environment and installs dependencies.

   ./install_zeta.sh

3. Start Zeta To launch both the Backend Agent and the Web Dashboard:

   zeta start

   Access the Dashboard at http://localhost:5173

Zeta's Intent Router understands context. Try these:

• Research: "Find the latest news on Artificial General Intelligence."
• Finance: "What is the price of Apple stock?"
• Memory: "My name is Pranav and I like coding." -> Later: "What is my name?"
• System: "List files in this directory." (Prompts for approval).

Edit config.yaml to customize:

• LLM Model: Change llm.model to any model available in Ollama.
• Risk Tolerance: Set safety.risk_tolerance to high to reduce confirmation prompts.

graph TD
    User[User] -->|Input| Router[Intent Router]
    Router -->|Chat| LLM[Ollama LLM]
    Router -->|Search| Web[Web Search Skill]
    Router -->|System| Sys[System Skill]
    
    LLM --> Memory[Memory Manager]
    Memory --> SQL[SQLite (Session History)]
    Memory --> Vector[ChromaDB (Long-term Facts)]
    
    API[FastAPI Server] -->|Exposes| Agent
    Dashboard[React Web UI] -->|Connects| API

Contributions are welcome! Please read CONTRIBUTING.md (coming soon) for details.

This project is licensed under the MIT License - see the LICENSE file for details.

Built with ❤️ by Pranav Dwivedi