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/cdavid817/vanehub-ai/agents-mdgit clone --depth 1 https://github.com/cdavid817/vanehub-aiWrote 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/cdavid817/vanehub-ai/agents-md)<a href="https://agentmods.dev/instructions/cdavid817/vanehub-ai/agents-md"><img src="https://agentmods.dev/badge/instructions/cdavid817/vanehub-ai/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 | $0.03067 | $0.03067 |
| Opus 5 | $0.01533 | $0.01533 |
| Sonnet 5 | $0.00613 | $0.00613 |
| Haiku 4.5 | $0.00307 | $0.00307 |
Grade A, and why
vanehub-ai 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 5d 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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
本文件是 VaneHub AI 项目所有 AI 编程助手(Claude Code / Gemini CLI / Antigravity CLI / OpenCode / Codex 等)的统一入口。 CLAUDE.md、GEMINI.md 均指向本文件,请勿分别维护三份不同内容。 详细技术选型说明与完整代码规范见
openspec/project.md;具体场景的实现示例见.claude/skills/。
项目概览
VaneHub AI 是一个桌面端多 AI 编程助手管理终端,用于统一管理和切换 Claude Code、OpenCode、Codex CLI、Gemini CLI、Antigravity CLI 等多个 AI 编程代理。同一套 React UI 既可运行在 Tauri 桌面客户端内,也可通过 Web/mock adapter 以浏览器页面形式运行。
技术栈(严格约束,不允许引入替代方案)
- 前端:React 19 + TypeScript(strict mode)+ Vite
- 桌面运行时:Tauri 2.x(Rust)
- 状态管理:仅用 React 内置 state/context,不引入 Redux/Zustand/MobX
- 样式:Tailwind CSS,不写内联 style,不引入 styled-components/CSS Modules/其他 UI 组件库
- 数据库:SQLite,通过 Rust 侧访问,前端不直接连库
- 测试:Vitest(单元/组件测试,CI 强制覆盖率门槛)+ Playwright(E2E)
- 包管理:npm(项目已有 package-lock.json,不要切到 pnpm/yarn)
架构核心约束
- React 组件必须依赖
src/services/agent-service.ts定义的服务接口,禁止组件内直接调用 Tauriinvoke() src/services/tauri-agent-client.ts(桌面实现)与src/services/web-agent-client.ts(Web/mock 实现)必须保持接口一致,新增能力要同时改两处src-tauri/负责 Rust 侧的 CLI 检测、启动路由、SQLite 注册表与会话状态,不要把这类逻辑下沉到前端
日志规范
- 新增诊断日志、操作日志、前端持久化日志事件时,必须遵循
openspec/specs/unified-log-management/spec.md - Rust/native 侧日志必须通过统一日志服务写入,禁止新增 feature-local 日志文件或绕过脱敏直接落盘
- SDK/CLI/任务类操作输出必须保留页面内展示能力,同时写入统一日志目录
- React 组件不得直接写本地日志文件;需要持久化的前端错误必须通过 service boundary 上报到 native logging service
- 日志必须支持
error、warn、info、debug级别语义,敏感信息必须在落盘前脱敏
代码规范(可执行规则,详细版见 openspec/project.md)
- 提交前必须跑通本文件末尾「校验命令」一节的全部命令。不要在这里另维护一份子集——两份清单迟早漂移,而漂移的那一半正是 CI 会拦下来的东西
- TypeScript:禁止
any,禁止// @ts-ignore(需要绕过时用// @ts-expect-error并写明原因) - React:函数组件 + Hooks,禁止 class component;单文件不超过 300 行
- Rust:跨 Tauri command 边界的错误必须转换为
Result<T, String>或自定义 error enum,unwrap()/expect()仅限测试代码 - 注释只写"为什么这样做",不写代码翻译式注释
项目文件规范(概要,完整版见 openspec/project.md)
src/
├─ components/ # 纯展示型 React 组件,不直接依赖 Tauri API
├─ services/ # 前端服务边界层(唯一允许被组件依赖的一层)
├─ hooks/ # 自定义 hook
src-tauri/
├─ src/commands/ # 每个 Tauri command 一个文件,按功能域分组
├─ src/platform/database/ # SQLite 访问层(schema 与迁移)
openspec/
├─ changes/ # 未归档的变更提案
│ └─ archive/ # 已完成变更的历史记录
├─ specs/ # 已确认规范(唯一真源)
└─ project.md # 项目上下文和详细规范
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.
- 5d ago First seen · 120 lines · 3,067 tokens per session scan A 195139c3391c
vanehub-ai AGENTS.md is an instructions file published in the GitHub repository cdavid817/vanehub-ai (110 stars, last pushed yesterday), licensed Apache-2.0. It adds 3,067 tokens to every session, about $0.0153 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
prompt-optimizer CLAUDE.md
Claude Code instructions for CalvinTTooS/prompt-optimizer, covering claude.md — prompt-optimizer, contesto del progetto, scelte architetturali vincolanti, organizzazione del codice and comandi del progetto (concreti ed eseguibili).
AgentDock AGENTS.md
Instructions for Cailiang/AgentDock, a project described as: Beginner-friendly desktop hub for AI coding clients, providers, Skills and MCP | 面向普通用户的 AI 编程客户端管理工具.
prompt-optimizer AGENTS.md
AGENTS.md instructions for CalvinTTooS/prompt-optimizer: This version has breaking changes — APIs, conventions, and file structure may all differ from your training data. Read the relevant guide in nodemodules/next/dist/docs/ before writing any code. Heed deprecation notices.
ClawRouter CLAUDE.md
Claude Code instructions for BlockRunAI/ClawRouter, covering clawrouter, commands, project structure, key dependencies and conventions.
InvestSkill GEMINI.md
Gemini CLI instructions for yennanliu/InvestSkill, covering investskill — gemini cli setup & usage guide, installation & setup, quick start, navigate to the investskill directory and start gemini cli (loads gemini.md automatically).
aisix CLAUDE.md
Claude Code instructions for api7/aisix, covering claude.md, 1. think before coding, 2. simplicity first, 3. surgical changes and 4. goal-driven execution.