Workspace

A personal workspace for brainstorming, planning, troubleshooting, and building — with AI collaboration conventions baked in.

This repository serves as a centralized hub for my technical projects, documentation, and experimentation. It uses a structured approach to collaboration with AI agents to maintain continuity and capture knowledge.

This workspace operates on an "upstream first" paradigm: rather than building custom, isolated solutions directly within client or downstream environments, we author, refine, and publish general-purpose examples, reusable artifacts, modules, and documentation open-source first. These upstream-validated assets are then pulled down into target environments as clean, standard deliverables, reducing technical debt and accelerating delivery across engagements.

Special thanks to hhellbusch/my-ai-workspace for providing much of the inspiration and foundational patterns for the AI collaboration workflows and workspace structure used here.

• Project planning and documentation — Architecture designs, decision logs, technical docs (OpenShift, Homelab, OpenBao, etc.)
• Work-in-progress sessions — Brainstorming, debugging investigations, and scratch work in wip/
• Reusable artifacts — Scripts, playbooks, and configs that can be applied across projects
• Persistent context — Session handoffs and state that survives across working sessions

.
├── .agents/                # Installed AI agent skills and behaviors
├── artifacts/              # Reusable scripts, playbooks, and configs
├── docs/                   # Finished documentation and guides
│   ├── homelab/            # Homelab infrastructure planning
│   └── troubleshooting/    # Solutions for common technical issues
├── planning/               # Session state, tracking, and roadmaps
├── research/               # Deep dives and comparative analysis
├── wip/                    # Active brainstorming and investigations
└── workstyle/              # AI collaboration conventions (Working Style)

This workspace enforces strict alignment and controlled execution through dedicated Skills and Conventions. These patterns help maintain continuity across sessions and capture knowledge as work happens.

To start a session with an AI assistant, ask it to load the conventions:

"Load workstyle/working_style.md as your working style, and validate the behaviors and conventions you now have."

See workstyle/working_style.md and AGENTS.md for the full conventions.