prompt
FreeEngineering policies for AI coding agents from classic software canon
About prompt
This is a binding engineering policy prompt for AI coding agents, derived from classic software engineering texts: Clean Code / Clean Architecture (Robert C. Martin), Domain-Driven Design Distilled (Vaughn Vernon), and Designing Data-Intensive Applications (Martin Kleppmann). It converts these principles into a set of rules that AI agents must follow when generating, editing, or reviewing code. The prompt emphasizes local reasoning, precise naming, inward dependencies, bounded contexts, explicit domain language, data ownership, consistency, and durability contracts. It uses MUST, SHOULD, and MUST NOT to indicate binding rules. Originally posted in the ai-boost/awesome-prompts GitHub repository.
Key Features
Pros & Cons
- Promotes highly readable and maintainable code
- Enforces strong architectural boundaries
- Reduces technical debt through consistent rules
- Improves collaboration between human developers and AI agents
- Based on proven software engineering literature
- May be overly strict for small or rapid prototyping projects
- Requires discipline to follow all binding rules
- Not a replacement for human code review or domain knowledge