Coding Agent Rules: Write Clean, Safe Code

Enforce strict coding standards for AI agents—prevent errors, maintain scope, and ensure production-ready quality with clear rules and checks.

14 views

0 saves

Workflow Steps

AI_RULES.md

> Coding agent behavior contract — read this before touching any file.

---

## 1. Identity & Role

You are a **senior full-stack engineer** embedded in this project. You write production-quality code, ask clarifying questions before making assumptions, and never over-engineer.

---

## 2. Core Behavioral Rules

ARCHITECTURE.md

> The single source of truth for how this system is built and why.

---

## 1. Project Overview

**Project name:** `[PROJECT_NAME]`
**Type:** `[Web App / API / CLI / Mobile / Monorepo]`
**Primary language(s):** `[TypeScript / Python / Go / ...]`
**Last updated:** `[YYYY-MM-DD]`

---

PLAN.md

# 📋 PLAN.md
> The execution roadmap. What we're building, in what order, and what done looks like.

---

## 1. Current Status

**Phase:** `[Discovery / Build / Polish / Launch / Post-launch]`
**Sprint / Week:** `[e.g. Week 3 of 6]`
**Last updated:** `[YYYY-MM-DD]`
**Updated by:** `[@handle]`

PRD.md

# 📄 PRD.md — Product Requirements Document
> What we're building, for whom, and how we know it's working.

---

## 1. Document Info

| Field | Value |
|---|---|
| **Product / Feature** | `[Name]` |
| **Author(s)** | `[@handle]` |
| **Status** | `[Draft / In Review / Approved / Shipped]` |

Coding