# Agent Rules Books — AGENTS.md from SE Books

> agent-rules-books ships MIT-licensed AGENTS.md/skills distilled from classic software engineering books, with mini/nano/full variants and size metrics.

## Install

Paste the prompt below into your AI tool:

# Agent Rules Books — AGENTS.md from SE Books

> agent-rules-books ships MIT-licensed AGENTS.md/skills distilled from classic software engineering books, with mini/nano/full variants and size metrics.

## Quick Use

1) Install

```bash
git clone https://github.com/ciembor/agent-rules-books.git
```

2) Run

```bash
Copy one `mini` rule set into AGENTS.md (or a Claude/Cursor rules file)
```

3) Verify

```bash
Ask your agent to restate the rules and apply them on a small refactor
```

---

## Intro

**Best for**: teams that want grounded engineering rules (refactoring, architecture, DDD) ready for Codex/Cursor/Claude Code

**Works with**: Any agent that reads Markdown rule files (AGENTS.md / CLAUDE.md / Cursor rules). Copy-based setup.

**Setup time**: 6 minutes

### Quant Data

- Each ruleset ships `full/mini/nano` variants with line/size metrics (repo)
- Setup time ~6 minutes

---

## How to Use It Well

Use these rules as a working agreement. Keep the active set small, and add a short checklist at the top: tests first, small diffs, stop early.

### Adoption Checklist

- Start with one real task and keep the scope narrow
- Capture a baseline: time-to-first-success and output quality
- Version your config/skills so teammates stay in sync

### Guardrails

Rules are only useful when enforced. Pair them with CI gates (tests, lint) and code review conventions.

### FAQ

**Q: Is this copyrighted book content?**
A: The README states the rules are inspired by books and intentionally avoid reproducing book text; treat them as operational instructions.

**Q: Which variant should I start with?**
A: Use `mini` for most tasks; keep `nano` for tight context budgets and `full` as reference.

**Q: How do I combine multiple books?**
A: Start with one baseline (e.g., Refactoring), then layer a second only after you see consistent behavior.

---

## Source & Thanks

> GitHub: https://github.com/ciembor/agent-rules-books
> Owner avatar: https://avatars.githubusercontent.com/u/702108?v=4
> License (SPDX): MIT
> GitHub stars (verified via `api.github.com/repos/ciembor/agent-rules-books`): 1,305


---

<!-- ZH -->

# Agent Rules Books——书籍提炼的 AGENTS.md 规则

> agent-rules-books 提供 MIT 许可的 AGENTS.md 规则/技能，来自经典软件工程书籍的可执行指令，可直接复制使用，并给出 mini/nano/full 三档与行数/大小指标。

## 快速使用

1）安装

```bash
git clone https://github.com/ciembor/agent-rules-books.git
```

2）运行

```bash
Copy one `mini` rule set into AGENTS.md (or a Claude/Cursor rules file)
```

3）验证

```bash
Ask your agent to restate the rules and apply them on a small refactor
```

---

## 简介

**适合谁**：希望把重构/架构/DDD 等工程规则变成可直接给 Codex/Cursor/Claude Code 用的团队

**适用环境**：支持读取 Markdown 规则文件的代理（AGENTS.md / CLAUDE.md / Cursor rules）；通过复制粘贴接入。

**安装耗时**：6 分钟

### 量化信息

- 每个规则集提供 full/mini/nano 三档，并标注行数/大小（仓库）
- 接入约 6 分钟

---

## 用好它的方式

把它当成团队 working agreement：活跃规则集保持足够小，并在顶部加一个短 checklist（先测试、小 diff、及时止步）。

### 推广清单

- 先选一个真实任务，小范围试跑
- 记录基线：首次成功耗时与输出质量
- 配置/技能要版本化，避免团队漂移

### 风险与护栏

规则只有在被执行时才有用：配合 CI（测试/lint）与 code review 约定一起落地。

### FAQ

**Q: 会不会包含书籍原文？**
A: README 说明这些规则受书籍启发，并刻意避免复述原文；把它当成工程操作指令使用。

**Q: 先用哪个版本？**
A: 大多数任务从 `mini` 开始；上下文很紧时用 `nano`；`full` 作为参考。

**Q: 多个书怎么叠加？**
A: 先选一个基线（如 Refactoring），观察行为稳定后再叠第二层，避免指令互相打架。

---

## 来源与感谢

> GitHub：https://github.com/ciembor/agent-rules-books
> Owner avatar：https://avatars.githubusercontent.com/u/702108?v=4
> 许可证（SPDX）：MIT
> GitHub stars（已通过 `api.github.com/repos/ciembor/agent-rules-books` 核验）：1,305


---
Source: https://tokrepo.com/en/workflows/agent-rules-books-agents-md-from-se-books
Author: Prompt Lab