Aegis Synthesis Architecture (ASA)

A fully-featured personal AI assistant with local LLM, secure P2P collaboration, proactive agents, and adaptive learning.

Features

• 🧠 Local LLM - Llama 3.2 or Mistral 7B with hot-swapping
• 🔐 Secure Mesh Network - E2EE P2P collaboration with consent tokens
• 🎯 Proactive Agents - Context-aware suggestions (Sentinel & Curator)
• 📚 RAG System - Vector knowledge base with semantic search
• 🔄 CRDT Memory Graph - Distributed fact synchronization
• 📖 Adaptive Learning - Style adaptation & LoRA fine-tuning
• 🛡️ Sandboxed Code Execution - Safe Python code execution

A privacy-first AI assistant that runs entirely on your computer - no cloud, no data sharing, no third-party APIs.

Core Features:

Local AI Brain: Multiple AI models you can switch between (small/fast or large/smart) Smart Memory: Remembers facts with your approval, searches past conversations Proactive Helpers: Background agents that suggest actions (clipboard monitoring, knowledge suggestions) Multi-Device Sync: Securely connect your own devices with end-to-end encryption Learning System: Gets better over time from your feedback Web Interface: Easy-to-use chat UI in your browser Key Design Principles:

Privacy: Everything stored locally in SQLite databases Control: You approve what it remembers Extensible: Tool system for calculations, web search, code execution Secure: Optional P2P collaboration uses military-grade encryption with explicit consent Reliable: Can run 24/7 in headless mode, packages as standalone app Architecture Highlights:

Async Python throughout for responsiveness ReAct reasoning loop for tool use CRDT for distributed memory sync across devices Consent tokens for collaboration security WAL-mode SQLite for data integrity Think of it as: Your personal ChatGPT that lives on your computer, remembers what you tell it, can connect your devices securely, and never sends your data anywhere.

Quick Start

# 1. Setup environment
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

# 1. Clone and setup
git clone https://github.com/Mr5niper/Aegis-Synthesis-Architecture/ aegis-synthesis
cd aegis-synthesis
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

# 2. Install dependencies
pip install -r requirements.txt

# 3. Run Aegis
python -m src.main_gui

First run downloads ~2GB model automatically (5-10 min). Browser opens to http://localhost:7860 when ready.

Try: "Hello" → "What's 23 * 456?" → "Remember my favorite color is blue"

---