# **Coding Guidelines**

## **Directive Category**

(Used when a statement contains a recommendation on how GenAI should be used or what to pay attention to.)

* **Reviewing Outputs** — Carefully check and test GenAI outputs for errors (do not trust or adopt them blindly).
* **Appropriate Context** — Provide the model with sufficient and appropriate input information (“prompt properly”).
* **Project-wide Consistency** — Ensure consistency across multiple outputs (e.g., coding conventions, design patterns, team-wide understandability).
* **Retain Understanding** — Avoid losing personal understanding; ensure continued learning.
* **Use Own Solutions** — Do not generate solutions directly; avoid asking the model for initial ideas or complete solutions.
* **Improving Solutions** — Use GenAI to refine or improve one’s own ideas or drafts.
* **Appropriate Degree of Usage** — Avoid excessive use; some statements advise limiting or generally cautioning against GenAI usage.
* **Careful Choice of Models** — Use the appropriate model or tool; includes concrete tool recommendations.

---

## **Use Case Category**

(Used when a statement describes a situation or task for which GenAI should be used.)

* **Documentation** — Support for documentation tasks (including code comments).
* **Learning** — Help with learning or understanding new technologies or frameworks.
* **High-Level Design** — Using GenAI to suggest processes, technologies, architectures, or designs.
* **User Interface** — Assistance in designing or implementing the UI.
* **Trivial Tasks** — Support for simple or routine tasks (e.g., basic or boilerplate code).
* **Debugging** — Identifying or fixing errors.