Start Here

[Varietyz]

Welcome to Disciplined AI Software Development

This repository is a methodology for working with AI on software projects. It is not a codebase. There is nothing to install, no dependencies, no build step. You read it, apply it to your own project, and the AI becomes a disciplined collaborator.

What this methodology covers

• Architectural principles that keep AI-generated code clean and maintainable
• Development phases that structure how work progresses from infrastructure through implementation
• Behavioral constraints that shape how the AI operates (CLAUDE.md policy, session governance)
• Validation hooks that catch what the AI misses (automated enforcement, feedback cycles)
• PAG (Pattern Abstract Grammar) for writing structured AI instructions
• claude-setup/ as a bootstrapping template you copy into your own project

Where to start reading

1. methodology/README.md — the front door. Section headings are the navigable index.
2. Browse the methodology documents as concepts come up in the narrative.
3. When ready to apply: copy claude-setup/ into your project and run the initialization.

How to interact

• Have a question? Open a Q&A Discussion or a Getting Started Discussion
• Applied it to your project? Share in Show & Tell
• Found a broken link or unclear section? Open a Content Issue
• Have an improvement idea? Open a Suggestion
• Want to discuss approaches? Join Methodology Chat

What AI tools does this work with?

The methodology works with any CLI-based AI tool. It works best with Claude Code because claude-setup, PAG, and the memory governance system are built for it. It also works with Cursor, Windsurf, and similar tools with adaptation.

Monthly snapshots

Each month a tagged release captures a stable point-in-time snapshot. Clone a release if you want a fixed reference that will not change under you.