⚙️
Setup
The workstation, cluster and local AI I operate
🌟 Workstation Main
Active
⚡
CPU
high-core desktop class
🎮
GPU
Local GPU compute
💾
Memory
Plenty for local models
📁
Storage
Fast local NVMe
🐧
OS
Omarchy (Arch Linux)
⚙️ Software Stack
🐧
OS
Omarchy (Arch)
✨
Editor
Claude Code with Serena semantic tools
🐚
Shell
Bash with custom functions
🦊
Browser
Firefox with privacy extensions
💬
Communication
GitHub, email
💡 Philosophy
📜
Declarative
Omarchy configuration as source of truth
🔄
Reproducible
Same config = same system, every time
📚
Documented
Comprehensive AGENTS.md for all AI assistants
🏠
Local-First
Self-hosted AI, minimal cloud dependencies
🔒
Privacy-Focused
Full disk encryption, VLANs, OPSEC
🔁 How the work gets done
The loop I run, on this hardware. Not a syllabus.
🤖
Agent delegation
Scoped tasks on role-specific profiles, dispatched from the kanban board
🔍
Review and merge
Read the diff, check the evidence, keep the previous generation for rollback
☸️
Cluster operations
A small k3s homelab declared in git and delivered by Argo CD
🔌
MCP tooling
Servers in front of the services so agents act on real systems
🧠
Local models
llama.cpp on owned GPUs, so no prompt leaves the cluster
🏠 Infrastructure
A small k3s homelab on Omarchy, with AI-powered monitoring:
- Multi-host Omarchy scripts deployment
- Real-time monitoring with VictoriaMetrics and Grafana
- Argo CD GitOps since September 2026
- 30+ documentation files
📁 Configuration
Full Omarchy configuration available on GitHub. Includes:
- Flakes-based system configuration
- Omarchy scripts multi-host deployment
- Custom modules for AI and monitoring
- Agent-friendly documentation