Project Aegis Documentation¶
Comprehensive architecture, operations, and disaster recovery documentation for Project Aegis - an autonomous AI agent platform.
Quick Links¶
| For... | Start Here |
|---|---|
| New Users | Getting Started |
| Operators | Operations |
| Developers | Development |
| Disaster Recovery | DR Runbook |
System Status¶
| Service | URL | Status |
|---|---|---|
| Dashboard | aegisagent.ai | Production |
| Intel | intel.aegisagent.ai | Production |
| Notebooks | notebooks.aegisagent.ai | Production |
| Code | code.aegisagent.ai | Production |
| VNC | vnc.aegisagent.ai | Production |
Documentation Structure¶
Getting Started¶
- Overview - What is Project Aegis
- Quick Start - 5-minute orientation
- First Commands - Essential commands
Architecture¶
- System Overview - High-level architecture
- Infrastructure - Hetzner, LXC, resources
- Network - Topology, DNS, SSL/TLS
- Containers - Docker stack, health checks
- Database - PostgreSQL schema (137 tables)
- Memory - Multi-tier memory architecture
- LLM Routing - Cognitive hierarchy
Services¶
- Dashboard - FastAPI application
- Scheduler - APScheduler jobs
- Traefik - Reverse proxy
- FalkorDB - Knowledge graph
- PostgreSQL - Database
- Playwright - Browser automation
Integrations¶
- Discord - Bot and channels
- Telegram - Alert channel
- WhatsApp - Vonage WABA
- Email - Gmail and Resend
- GitHub - Repository management
- Stripe - Payment processing
- Google Workspace - Gmail, Calendar
MCP Servers¶
- Configuration - Setup and troubleshooting
- Available Servers - All MCP integrations
- Creating Servers - Development guide
Operations¶
- Daily Routines - Morning, evening, maintenance
- Monitoring - Health checks, alerts
- Backup & Recovery - Procedures
- Troubleshooting - Common issues
Security¶
- Credentials - Storage and rotation
- Access Control - API keys, permissions
- Hooks - Pre/post tool execution
Development¶
- Module Structure - 78 Python modules
- Testing - pytest patterns
- Contributing - Git workflow
Products¶
- Intel Dashboard - Geopolitical intelligence
- Research API - REST API
- Open Notebook - Research tool
Reference¶
- API Reference - All endpoints
- CLI Commands - Slash commands
- Configuration - Settings files
- Environment Variables - Complete list
Disaster Recovery¶
Key Concepts¶
Cognitive Hierarchy¶
| Tier | Model | Use Case |
|---|---|---|
| 1 | Claude Opus 4.5 | Strategic decisions (rare) |
| 1.5 | Claude Haiku 4.5 | Fast ops (classify, extract) |
| 2 | GLM-4.7 via Z.ai | 90%+ operational work |
| 3 | Ollama Local | Fallback, offline, vision |
Work Priorities¶
- SHIP & ACTIVATE (70%) - Wire existing features
- REACTIVE WORK (20%) - Discord tasks, Beads issues
- PROACTIVE WORK (10%) - Research, documentation
Memory Architecture¶
- Episodic: Events, decisions, interactions
- Semantic: Knowledge, facts, learnings
- Procedural: Workflows, how-tos
- Knowledge Graph: FalkorDB entity relationships
Emergency Contacts¶
- Discord: Guild 1454722052777836546, #alerts channel
- WhatsApp: +447441443388
- Telegram: Chat ID 1275129801
- Email: aegis@richardbankole.com
Last Updated: 2026-01-25 Version: 1.0 Source: GitHub