LAIBRARY
Aircury's AI Library
Where we document what we know about building software with AI. How we work, what we've learned, and the thinking behind the decisions we make day to day.
How we build
The Framework
How we approach software development at Aircury — using OpenSpec as base and adding the architectural habits that stop AI from producing code that works today but becomes a problem in six months.
What we know
AI Engineering
The concepts and mental models that help you work with AI systems effectively — not just using them, but understanding what they actually do, where they tend to fail, and how to build reliable things on top of them.
Quick Reference
Methodology
The OpenSpec Extended cycle and the spec-to-test idea
Architecture
DDD + Hexagonal — why it matters when AI writes your code
Testing Strategy
The testing pyramid and the rebuildable codebase
SOLID Principles
Five principles that shape AI output towards maintainable design
Prompt Engineering
How to write prompts that produce predictable results
Evaluation
How to know if an AI-based system is actually working well
AI Agents
Agentic architectures and how to design them with rigor
RAG
The full Retrieval-Augmented Generation pipeline