An open engineering laboratory for AI-assisted software development.
Most AI coding resources focus on prompts, tools or model comparisons.
This repository focuses on something different:
engineering practices.
How should software engineers and AI coding agents work together to build reliable software?
This repository captures reusable engineering knowledge through practical engineering, real-world experience and continuous refinement.
Every document should answer a real engineering question. If it doesn't help solve a practical engineering problem, it probably doesn't belong here.
If you're new to the repository, begin with the Knowledge section.
Knowledge articles introduce the engineering principles that the rest of the repository builds upon.
Playbooks, policies and templates evolve from these foundational concepts.
AI coding is evolving faster than engineering practices.
Every week brings new:
- models;
- coding agents;
- frameworks;
- workflows.
But much of the shared knowledge still revolves around:
- prompts;
- tool comparisons;
- isolated examples;
- viral posts.
Good engineering practices don't appear overnight.
They emerge from building real software, solving real problems and continuously refining the way we work.
Instead of asking:
"What prompt should I use?"
we ask:
"What engineering workflow should I follow?"
Every document in this repository starts with a real engineering problem.
Real Project
↓
Engineering Problem
↓
Engineering Decision
↓
Generalization
↓
Engineering Principle
↓
Knowledge
↓
Playbook
↓
Policy
↓
Automation
The goal is not to document a specific project.
The goal is to extract reusable engineering knowledge that can be applied across different products, teams and technology stacks.
Documentation about this repository itself.
Examples:
- Writing Guidelines
- Repository Structure
- Contribution Guide
- Style Guide
- Roadmap
Engineering concepts, principles and patterns.
Examples:
- Context Engineering
- Product Workflows
- Feature Specifications
- Planning
- Memory
- MCP
- Tool Calling
- Multi-Agent Systems
🌱 RFCs
Ideas that are still evolving.
RFCs are proposals for engineering practices, workflows or policies.
They exist to collect discussion and feedback before becoming part of the knowledge base.
Practical workflows you can apply immediately.
Examples:
- Planning a Feature
- Architecture Review
- Refactoring
- Code Review
- Bug Investigation
- Starting a New Project
📋 Policies
Decision rules for humans and AI agents.
Examples:
- When should planning happen?
- When should an agent ask for clarification?
- When is an ADR required?
- When should context be expanded?
- When is human approval required?
Reusable engineering artifacts.
Examples:
- Product Workflow
- Feature Specification
- ADR
- Task Breakdown
- Planning Prompt
- Review Prompt
- Issue Template
This repository is designed to be useful for both:
- software engineers learning modern AI-assisted development;
- AI coding agents consuming structured engineering knowledge.
As the repository evolves, its structure should become increasingly useful for both humans and software.
This is intentionally not:
- a prompt collection;
- an awesome list;
- documentation for one AI tool;
- a benchmark leaderboard.
It is an evolving engineering knowledge base.
Some ideas will be wrong.
Some workflows will evolve.
Some recommendations will eventually be replaced by better ones.
That's expected.
Engineering improves through iteration.
So should engineering knowledge.
- Developers learning Agentic Engineering
- Senior engineers adopting AI workflows
- Engineering teams defining internal practices
- Tool builders
- AI agent developers
- Anyone who wants to move beyond vibe coding
Contributions are welcome.
The best engineering practices emerge through shared experience.
If you've developed a workflow, discovered a useful pattern or learned something while building software with AI, we'd love to hear about it.
Useful contributions include:
- reusable workflows;
- engineering patterns;
- playbooks;
- real-world examples;
- design decisions;
- lessons learned;
- thoughtful discussions.
The repository is still in its early stages.
Current priorities include:
- foundational engineering concepts;
- reusable workflows;
- practical playbooks;
- structured engineering knowledge;
- policy-driven development.
Long term, this repository aims to become a shared engineering handbook for the AI era.
A place where practical workflows, patterns and engineering decisions continuously evolve through real-world use and community discussion.