██████╗ ██╗ ██╗███████╗███████╗
██╔═████╗╚██╗██╔╝██╔════╝██╔════╝
██║██╔██║ ╚███╔╝ █████╗ █████╗
████╔╝██║ ██╔██╗ ██╔══╝ ██╔══╝
╚██████╔╝██╔╝ ██╗██║ ██║
╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═╝
AI Engineering Operating System
Engineering principles, decision frameworks, and behavioral rules that teach AI assistants to collaborate like experienced software engineers.
- Status: Active Development
- Version: v0.1.0
- License: MIT
AI assistants often overengineer simple work, change code before understanding it, or take control away from the developer. 0xff-AuDHD gives them a practical engineering standard: understand first, keep changes focused, and prefer simple solutions with clear value.
- Developers
- Students
- Open source contributors
- Freelancers
- Anyone tired of AI overengineering simple problems
0xff-AuDHD helps you give an AI assistant a consistent way to approach engineering work.
An AI Engineering Operating System is a shared set of rules for how an AI should plan, change, explain, and review software. It keeps the developer as the project owner and the AI as a collaborator.
Why: This downloads 0xff-AuDHD to your computer.
git clone https://github.com/sumit0xff/0xff-AuDHD.git
cd 0xff-AuDHDWhy: SKILL.md defines the engineering behavior for your AI assistant.
cat SKILL.mdWhy: Provide the skill and relevant project context before requesting work.
See the platform-neutral Usage Guide.
| Platform | Status |
|---|---|
| ChatGPT | Supported |
| Claude | Supported |
| Codex | Supported |
| Cursor | Supported |
| Gemini | Supported |
| Antigravity | Supported |
Platform integrations may differ, but the engineering philosophy remains the same.
| Path | Responsibility |
|---|---|
SKILL.md |
Executable behavior specification for AI assistants. |
docs/ |
Philosophy, usage, installation, and roadmap documentation. |
rules/ |
Rules that support the philosophy and skill. |
examples/ |
Focused examples of applying 0xff-AuDHD. |
tests/ |
Validation artifacts for documented behavior. |
integrations/ |
Materials for supported AI platforms. |
.github/ |
GitHub-specific repository configuration. |
CONTRIBUTING.md |
Contribution requirements and standards. |
CHANGELOG.md |
Version history. |
The philosophy is the foundation of every behavior in this repository. Read the complete Philosophy.
View the four laws
- Understand Before Changing — Understand the existing implementation before changing it.
- Minimize Unnecessary Complexity — Prefer simple solutions with clear value.
- Complexity Must Be Earned — Add complexity only when the software truly requires it.
- Teach Through Building — Teach while solving the current problem.
Task: Add email validation to an existing signup form.
“I’ll add a validation service, shared schema layer, reusable form hooks, and several new files.”
“I’ll inspect the existing signup validation first. If a suitable function exists, I’ll add the smallest email check there, test the form, and only extract shared code if it is needed.”
| Version | Focus | Status |
|---|---|---|
| v0.1 | Foundation: philosophy, SKILL, and documentation | Complete |
| v0.2 | Decision Engine, AI Smells, Context Detection, and Framework Rules | Planned |
| v0.3 | Integrations, benchmarks, validation, and community contributions | Planned |
Is this a prompt?
No. It is a set of engineering principles, decision frameworks, and behavioral rules.
Is this only for AuDHD?
No. It is designed for developers using the supported AI platforms.
Can I use it with ChatGPT?
Yes. Follow the shared workflow in the Usage Guide.
Can I contribute?
Yes. Read CONTRIBUTING.md before contributing.
Why is it open source?
The project is public so its philosophy, rules, and direction can be inspected and improved through contributions.
Read CONTRIBUTING.md before contributing.
This project is licensed under the MIT License.
Built for developers who value clarity over complexity.