Home / robozephyr / ai-engineering-playbook · AGENTS.md · GitHub

AGENTS.md agentsA

Name: AGENTS.md
Author: robozephyr

AGENTS.md is agent-read markdown (agents) from robozephyr/ai-engineering-playbook.

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Agents 指南

本文件约束所有 AI Agent 在此项目中的行为。

## 项目性质

这是一个 **文档站 / 知识库**，用于管理、评估和展示 AI 编码 Skills 与 Workflows。不是可执行的 Skill 库，不是 prompt 模板库。

## 内容生产规则

1. **只做评测和指导，不做模板输出**
   - 正确：分析某 Skill 的四维设计能力和反模式检测
   - 错误：输出"将以下 prompt 嵌入你的 CLAUDE.md"

2. **保持技术中立**
   - 不绑定特定框架（Vue/Nuxt/React/Angular）
   - Skill 评测应说明适用的技术栈范围，但不偏向任何一个

3. **信息不重复**
   - 每个 Skill 的详细评测只在对应领域页中出现一次
   - 其他页面通过链接引用
   - Workflow 页面引用 Skill 时链接到领域页，不重复描述

4. **评测结构统一**
   - 每个 Skill 评测包含：来源、安装方式、核心能力、我们的评价、最佳使用场景、注意事项、推荐搭配
   - 使用三级优先级：必装 / 推荐 / 视情况

5. **Workflow 结构统一**
   - 按 Phase 组织，每个 Phase 有步骤表格
   - 包含前置准备（安装命令）和快速参考（一行版）
   - 明确标注平台限制（如 Web-only、iOS-only）

## 文件组织（三层定位）

| 层 | 角色 | 路径 |
|---|---|---|
| **L1 方法论** | 怎么做 | `harness/` / `visual-engineering/` / `skills/index.md`（评测方法） |
| **L2 资产** | 用什么 | `skills/<领域>.md`（评测） / `workflows/<场景>.md`（编排） |
| **L3 Reference** | 看到了什么 | `research/<主题>-YYYY-MM.md`（带日期事实快照） |

```
harness/             — Agent 工作环境方法论（L1）
visual-engineering/  — AI 视觉生成方法论（L1）
skills/<领域>.md     — 按领域分类的 Skill 评测（L2）
skills/index.md      — Skills 知识中枢：概念 / 选型 / 评测方法 / 索引（L1+L2 入口）
workflows/<场景>.md  — 按场景编排的 Workflow（L2）
research/<主题>.md   — 带日期的事实快照（L3）
…

Read the whole file at its exact version.

How to install

Latest version

mdr add robozephyr/ai-engineering-playbook/AGENTS.md@git:20260519.2b008ba

Exact content

mdr add robozephyr/ai-engineering-playbook/AGENTS.md@sha256:dc1293958faba7e1

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

[![mdr](https://markdownregistry.com/badge/art_q3ttqiqlnhx27rmv.svg)](https://markdownregistry.com/a/art_q3ttqiqlnhx27rmv)

1 badge views in 30 days

Versions

version	committed	commit	size	audit
git:20260519.2b008ba latest	2026-05-19	2b008ba	2,260 B	A	view

Audit of the latest version

A 14 of 14 checks passed. Deterministic, no model, same answer every run.

pass: Size between 200 bytes and 200 KB (2260 bytes)
pass: No zero-width or bidi control characters
pass: No instruction hidden inside an HTML comment
pass: No link to an exfiltration or paste host
pass: No credential-shaped string
pass: No instruction to send local credentials anywhere
pass: No text hidden with inline styles
pass: No prompt-injection phrasing
pass: No curl or wget piped into a shell
pass: No recursive delete of root, home or parent
pass: No instruction to read or print local credentials
pass: No base64 blob over 200 characters
pass: No link to a raw IP address
pass: No script tag

Source

GitHub

robozephyr/ai-engineering-playbook · 3 stars · license none · pushed 2026-09-23 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_q3ttqiqlnhx27rmv
GET https://markdownregistry.com/api/v1/resolve?ref=robozephyr/ai-engineering-playbook/AGENTS.md
GET https://markdownregistry.com/api/v1/blob/dc1293958faba7e158eb63c2c6e791d297a506f613f130699e4a8637672b15e9

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.