Disciplined software engineering for the AI era.
AI has collapsed the distance between intent and implementation. A developer can describe a feature and get working code in minutes. A founder can assemble a prototype in a weekend. A team can ship more output than ever before. But this has brought about a new set of challenges. You have: more features than clarity, more code than comprehension, and more velocity than direction.
The AI SDLC helps teams use AI to build software without losing control, context, or understanding.
Important
The AI SDLC structures development around explicit control points — moments where the developer reviews what was built, understands it, and decides whether to advance. AI handles the implementation between those points. The developer owns the decisions, and decisions are preserved as shared context in the repo.
→ AI SDLC in Practice — phases, verification surfaces, and the responsibility split in detail
- Manifesto — read the manifesto
- Principles — learn the key principles
- In Practice — see the workflow
- Contributing — how to engage
Note
The tools changed. The job didn't. Support the AI SDLC manifesto by starring this repo.