# agents-cli — Agent Build/Eval/Deploy Skills for Coders

> agents-cli installs a CLI + skills so your coding assistant can scaffold, evaluate, and deploy production agents on Google Cloud with repeatable commands.

## Install

Save as a script file and run:

## Quick Use

1. Install (recommended):
   ```bash
   # Prereqs: Python 3.11+, uv, Node.js
   uvx google-agents-cli setup
   ```
2. Or install only the skills into your coding agent:
   ```bash
   npx skills add google/agents-cli
   ```
3. Confirm the CLI runs:
   ```bash
   agents-cli info
   ```

## Intro

agents-cli installs a CLI + skills so your coding assistant can scaffold, evaluate, and deploy production agents on Google Cloud with repeatable commands.

- **Best for:** teams building agents on Google Cloud that want repeatable scaffolding, evals, and deployment flows
- **Works with:** Python 3.11+, uv/uvx, Node.js; can be used standalone or from coding agents
- **Setup time:** 15–30 minutes

## Practical Notes

- Start local-first: scaffold + run + eval before touching cloud deploy.
- Keep your agent lifecycle reproducible by pinning tool versions in CI.

## Mental model: it’s “skills + lifecycle commands”

agents-cli is not a coding agent itself. It’s a set of CLI commands and skills that teach your coding assistant how to:
- scaffold an agent project,
- run and evaluate it,
- deploy and operate it on Google Cloud.

## Practical workflow (minimal)

### 1) Scaffold a project

```bash
agents-cli scaffold my-agent
cd my-agent
```

### 2) Run the agent locally (one-shot prompt)

```bash
agents-cli run "Summarize the latest logs and propose fixes"
```

### 3) Add evaluation early

Even if your eval set is small, it prevents regressions:

```bash
agents-cli eval run
```

## Team hygiene

- Treat the CLI as the “source of truth” for commands your agent should use.
- Keep a short `AGENTS.md` in the repo that says which commands are required before PR.
- When the agent changes code, ask it to run `agents-cli lint` (or your repo’s lint) before it claims done.

### FAQ

**Q: Is this a replacement for Claude Code/Codex/Gemini CLI?**
A: No. It augments them with skills and lifecycle commands for agent projects.

**Q: Can I use it without a coding agent?**
A: Yes. The CLI is designed to work standalone from your terminal.

**Q: Do I need Google Cloud for local development?**
A: Not for basic scaffold/run/eval; cloud is mainly for deployment and ops features.

## Source & Thanks

> Source: https://github.com/google/agents-cli
> License: Apache-2.0
> GitHub stars: 2,285 · forks: 262

---

<!-- ZH -->

## 快速使用

1. 安装（推荐）：
   ```bash
   # 前置：Python 3.11+、uv、Node.js
   uvx google-agents-cli setup
   ```
2. 或者只把 skills 安装到你的 coding agent：
   ```bash
   npx skills add google/agents-cli
   ```
3. 验证 CLI 可用：
   ```bash
   agents-cli info
   ```

## 简介

agents-cli 提供 CLI 与配套技能，让你的编码助手用可复用命令完成 Agent 项目脚手架、评测与部署到 Google Cloud 的全流程，并把关键步骤沉淀为团队标准与流水线。

- **适合谁：** 在 Google Cloud 上做 Agent 的团队，想要可复用的脚手架、评测与部署流程
- **可搭配：** Python 3.11+、uv/uvx、Node.js；既可独立使用，也可供 coding agent 调用
- **准备时间：** 15–30 分钟

## 实战建议

- 先本地闭环：scaffold → run → eval，再上云部署。
- 把版本固定进 CI，让 Agent 生命周期可复现、可回滚。

## 心智模型：它是“技能 + 生命周期命令集”

agents-cli 本身不是 coding agent；它是一组 CLI 命令与配套技能，用来教你的编码助手如何：
- 生成/升级 Agent 项目脚手架，
- 本地运行与评测，
- 在 Google Cloud 上部署与运维。

## 最小实战流程

### 1）创建项目

```bash
agents-cli scaffold my-agent
cd my-agent
```

### 2）本地运行（一次性 prompt）

```bash
agents-cli run "Summarize the latest logs and propose fixes"
```

### 3）尽早引入评测

哪怕 evalset 很小，也能显著降低回归：

```bash
agents-cli eval run
```

## 团队落地建议

- 把 CLI 当成“你的 agent 应该执行哪些命令”的权威来源。  
- 在仓库放一份精简 `AGENTS.md`，规定 PR 前必跑哪些命令。  
- 让 agent 改完代码后先跑 `agents-cli lint`（或你们仓库的 lint）再交付。

### FAQ

**它会替代 Claude Code/Codex/Gemini CLI 吗？**
答：不会。它是给这些工具“加技能 + 生命周期命令”的增强层。

**不用 coding agent 也能用吗？**
答：可以。CLI 支持你直接在终端独立运行。

**本地开发必须要 Google Cloud 吗？**
答：不必。scaffold/run/eval 可本地完成；云主要用于部署与运维。

## 来源与感谢

> Source: https://github.com/google/agents-cli
> License: Apache-2.0
> GitHub stars: 2,285 · forks: 262


---
Source: https://tokrepo.com/en/workflows/agents-cli-agent-build-eval-deploy-skills-for-coders
Author: Agent Toolkit