Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/southliu/south-admin-react/agents-mdgit clone --depth 1 https://github.com/southliu/south-admin-reactWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/southliu/south-admin-react/agents-md)<a href="https://agentmods.dev/instructions/southliu/south-admin-react/agents-md"><img src="https://agentmods.dev/badge/instructions/southliu/south-admin-react/agents-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01534 | $0.01534 |
| Opus 5 | $0.00767 | $0.00767 |
| Sonnet 5 | $0.00307 | $0.00307 |
| Haiku 4.5 | $0.00153 | $0.00153 |
Grade A, and why
south-admin-react AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
中后台管理前端项目:React 19 + TypeScript + Vite 7 + Ant Design 6 + Zustand,pnpm monorepo。详细说明可参考 CLAUDE.md 与 README.md。
常用命令
pnpm dev # 启动开发服务器
pnpm build # 生产构建(tsc + vite build)
pnpm lint # ESLint 检查并自动修复
pnpm lint:stylelint # 样式检查
pnpm prettier # 格式化代码
- 安装依赖必须加
-w(monorepo 根):pnpm i <package> -w;pnpm 10+ 构建脚本被拦截时执行pnpm approve-builds。 - 无测试框架,验证方式为
pnpm lint+pnpm dev/pnpm build。 - Git 提交遵循 conventional commits(husky + commitlint 强制,如
feat: xxx);提交失败先执行npx husky install。
新建页面(重要约定)
-
pages 目录按菜单结构命名:一级菜单 → 父文件夹,二级菜单 → 子文件夹(kebab-case)。如菜单为「系统管理-菜单管理、系统管理-用户管理、商城管理-商品类型、商城管理-商品列表」,对应:
src/pages/system/menu/index.tsx # 系统管理-菜单管理 src/pages/system/user/index.tsx # 系统管理-用户管理 src/pages/mall/goods-type/index.tsx # 商城管理-商品类型 src/pages/mall/goods/index.tsx # 商城管理-商品列表即
src/pages/<一级菜单>/<二级菜单>/index.tsx;同菜单层级的页面必须落在同一父文件夹下,路由/system/menu由目录自动生成。 -
优先使用
/demo-createskill(.claude/skills/demo-create/SKILL.md)生成标准 CRUD 页面,不要手写整套模板。 -
【强制】每次创建新的路由页面,必须使用
/demo-createskill 生成(.claude/skills/demo-create/SKILL.md),不要手写整套 CRUD 模板。仅以下情况可以不用 demo-create:页面非标准 CRUD 结构(如仪表盘、自定义复杂布局),此时在 AGENTS.md 或对话中说明原因。 -
每页固定
index.tsx + model.ts(model 承放搜索项、表格列、表单配置),文件夹划分要清晰分明。 -
能组件化就组件化:
- 仅当前页面使用的组件 → 放在该页面子文件夹的
components/中(参考src/pages/system/menu/components/)。 - 全局/跨页面使用的组件 → 放
src/components/,公共二次封装组件以Base开头(如BaseTable、BaseForm、BaseModal)。
- 仅当前页面使用的组件 → 放在该页面子文件夹的
-
接口文件放
src/servers/,目录结构必须与src/pages/保持一致:如src/pages/system/user/对应src/servers/system/user.ts。 -
样式尽量使用 UnoCSS 原子类,避免写冗余 less。
目录结构
src/pages/— 页面(路由按目录结构自动生成,见下)src/servers/— API 接口,结构与 pages 一致src/components/— 全局公共组件(Base 前缀)src/stores/— Zustand 状态(user/menu/tabs/public)src/hooks/useCommonStore.ts— 聚合各 store(useShallow优化),组件中优先用它而非直接访问 storesrc/locales/{zh,en}/— i18n 翻译(react-i18next)packages/— monorepo 子包:@south/request、@south/message、@south/utils、@south/stylelint
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 83 lines · 1,534 tokens per session scan A 15c039da1327
south-admin-react AGENTS.md is an instructions file published in the GitHub repository southliu/south-admin-react (578 stars, last pushed 13d ago), licensed MIT. It adds 1,534 tokens to every session, about $0.0077 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
ant-design-pro CLAUDE.md
Claude Code instructions for ant-design/ant-design-pro, covering claude.md, project, commands, critical rules and architecture essentials.
v3-admin-vite AGENTS.md
Instructions for un-pany/v3-admin-vite, covering 原则, 1. 编码前思考, 2. 简洁优先, 3. 精准修改 and 4. 目标驱动执行.
skyroc AGENTS.md
Instructions for Ohh-889/skyroc, covering agents.md, 规范索引, skills 调用规则(关键), 思考顺序 and react 组件规范.
pinyin AGENTS.md
Instructions for yunsii/pinyin, covering agents.md, hard rules, 方案与文本靠 side-effect 注册,漏 import 等于没加, vite.config.ts 的 base: '/pinyin/' 不要随手改 and 拼音术语变量名用全拼.
ant-design-pro AGENTS.md
AGENTS.md instructions for ant-design/ant-design-pro, a project described as: 👨🏻💻👩🏻💻 Use Ant Design like a Pro!
v3-admin-vite CLAUDE.md
Instructions for un-pany/v3-admin-vite, a project described as: ☀️ AI-friendly Vue3 admin template | Vue Admin | Vue Template | Vue3 Admin | Vue3 Template | Vue 后台 | Vue 模板 | Vue3 后台 | Vue3 模板.