Can You Vibe Code an OS Without Lying to Yourself?

Описание и применение

Автор описывает риски использования ИИ для генерации кода ОС и демонстрирует, как строгие инварианты и реальные тесты запуска помогают избежать критических ошибок.

Задача

Relying on AI-generated code for critical OS components can lead to undetected failures because tests may pass while the system is fundamentally incorrect.

Результат

By enforcing hard invariants and performing real boot tests, the author prevented a bricked update that would have passed all other checks.

Рабочий процесс

• Define critical invariants for OS components (e.g., update path must boot).
• Use AI agents to generate code against these invariants.
• Run automated tests covering metadata, signatures, and functional behavior.
• Perform a real boot of the updated system to validate the invariant.
• If the boot fails, investigate and reject the AI-generated code.
• Iterate until all invariants are satisfied and the system boots correctly.

Доказательства

The generated system was wrong. The architecture review did not catch it. The metadata tests did not catch it. Actually booting the update caught it.

Источник

Vibe coding an OS is a terrible idea if "vibe coding" means trusting generated code because it looks right and the tests are green.

Источник