{"id":477983,"date":"2026-04-29T14:41:06","date_gmt":"2026-04-29T14:41:06","guid":{"rendered":"https:\/\/savepearlharbor.com\/?p=477983"},"modified":"-0001-11-30T00:00:00","modified_gmt":"-0001-11-29T21:00:00","slug":"","status":"publish","type":"post","link":"https:\/\/savepearlharbor.com\/?p=477983","title":{"rendered":"Google Agents CLI: Full Beginner Guide (Step-by-Step)"},"content":{"rendered":"<div xmlns=\"http:\/\/www.w3.org\/1999\/xhtml\">\n<figure class=\"\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/974\/c7d\/f08\/974c7df08b8be52a8020845e34b02cf1.png\" alt=\"Google Agents CLI Explained in 5 Minutes\" sizes=\"(max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/\/post_images\/974\/c7d\/f08\/974c7df08b8be52a8020845e34b02cf1.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/974\/c7d\/f08\/974c7df08b8be52a8020845e34b02cf1.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/p>\n<div><figcaption>Google Agents CLI Explained in 5 Minutes<\/figcaption><\/div>\n<\/figure>\n<p>Google recently released the\u00a0<a href=\"https:\/\/google.github.io\/agents-cli\/\" rel=\"noopener noreferrer nofollow\"><strong>Agentic CLI<\/strong><\/a>, a powerful tool that makes building, testing, and deploying AI agents faster and more intuitive. I think it\u2019s super useful, so I\u2019ll walk you through the entire lifecycle of an AI agent using the CLI.<\/p>\n<h3>What is the Google Agents CLI?<\/h3>\n<p>Before we dive into the code, let\u2019s understand the \u201cwhy.\u201d Usually, building an AI agent requires a lot of \u201cglue code\u201d, the messy scripts that connect an LLM (Large Language Model) to external tools or databases.<\/p>\n<figure class=\"\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/630\/c3c\/ddb\/630c3cddb9d6d3ea82fee828f95e29a6.png\" alt=\"What is the Google Agents CLI\" sizes=\"(max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/\/post_images\/630\/c3c\/ddb\/630c3cddb9d6d3ea82fee828f95e29a6.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/630\/c3c\/ddb\/630c3cddb9d6d3ea82fee828f95e29a6.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/p>\n<div><figcaption>What is the Google Agents CLI<\/figcaption><\/div>\n<\/figure>\n<p>The Agents CLI removes this friction. Agents CLI is an official Google tool for creating, evaluating, and deploying agents built with Google\u2019s Agent Development Kit (ADK). It acts as the programmatic backbone of the Agent Development Lifecycle (ADLC) on Google Cloud.<\/p>\n<h4>Skills packaged for agents<\/h4>\n<p>Agents CLI bundles seven \u201cskills\u201d that teach coding agents how to perform each step of the ADLC:<\/p>\n<figure class=\"\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/0a3\/887\/50c\/0a388750c335351a5b0847ef2d50926d.png\" alt=\"Skills packaged for agents\" sizes=\"(max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/\/post_images\/0a3\/887\/50c\/0a388750c335351a5b0847ef2d50926d.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/0a3\/887\/50c\/0a388750c335351a5b0847ef2d50926d.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/p>\n<div><figcaption>Skills packaged for agents<\/figcaption><\/div>\n<\/figure>\n<p>Your coding agent can call these skills directly once installed. Alternatively, a human developer can run the same commands manually (see \u201cHuman Mode\u201d below).<\/p>\n<h3>How to Install Google Agents CLI?<\/h3>\n<p>To get started, you need a clean environment. The CLI relies on a few modern web and data tools. Follow these steps to ensure everything is ready:<\/p>\n<h4>Prerequisites<\/h4>\n<ul>\n<li>\n<p><strong>Python &amp; PIP:<\/strong>\u00a0The backbone of most AI development.<\/p>\n<\/li>\n<\/ul>\n<pre><code class=\"bash\">sudo apt install python3 python3-pip<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:87px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<ul>\n<li>\n<p><strong>Node.js &amp; NPM:<\/strong>\u00a0Required for the CLI\u2019s interface and deployment features.<\/p>\n<\/li>\n<\/ul>\n<pre><code class=\"bash\">sudo apt install nodejs npm<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<ul>\n<li>\n<p><strong>UV:<\/strong>\u00a0A lightning-fast Python package installer that the Agents CLI uses under the hood.<\/p>\n<\/li>\n<\/ul>\n<pre><code class=\"bash\">pip install uv<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<h4>Installation Steps<\/h4>\n<p>Create a Virtual Environment: It is always a good idea to keep your projects separate so they don\u2019t interfere with each other.<\/p>\n<pre><code class=\"bash\">python -m venv agent-env  source agent-env\/bin\/activate  # On Windows use `agent-env\\Scripts\\activate`<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>The CLI is distributed through\u00a0<a href=\"https:\/\/docs.astral.sh\/uv\/\" rel=\"noopener noreferrer nofollow\">uv<\/a>:<\/p>\n<pre><code class=\"bash\">uvx google-agents-cli setup<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>The CLI supports macOS and Linux; native Windows is currently unsupported (use WSL 2 instead).<\/p>\n<h4>Authentication<\/h4>\n<p>Agents CLI picks up your credentials automatically if you are already authenticated with the\u00a0<strong>gcloud<\/strong>\u00a0CLI. When running locally without gcloud, you can set a\u00a0<strong>Gemini API key<\/strong>:<\/p>\n<pre><code class=\"bash\">export GEMINI_API_KEY=\"your\u2011key\"<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>For detailed authentication scenarios (service accounts, A2A roles etc.) refer to the\u00a0<a href=\"https:\/\/google.github.io\/agents-cli\/guide\/authentication\/\" rel=\"noopener noreferrer nofollow\">Authentication guide<\/a>. The important point is that your coding agent inherits whatever credentials your shell has; you do not need to embed secrets in code.<\/p>\n<h3>Building Your First \u201cWeather\u201d Agent<\/h3>\n<p>The fastest way to learn is by doing. The CLI comes with a \u201cboilerplate\u201d or template system to get you moving in seconds.<\/p>\n<p>Run the following command in your terminal:<\/p>\n<pre><code class=\"bash\">agent-cli create my-first-agent  cd my-first-agent<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<h4>Install Dependencies<\/h4>\n<p>Inside your new project folder, you need to install the specific libraries required for this agent:<\/p>\n<pre><code class=\"bash\">agent-cli install<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<h4>The Playground<\/h4>\n<p>One of the best features of this CLI is the Playground. Instead of testing your agent in a black-and-white terminal, you can launch a local web interface.<\/p>\n<pre><code class=\"bash\">agent-cli playground<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>Once you run this, you\u2019ll get a local URL (usually\u00a0<code>http:\/\/127.0.0.1:8000<\/code>). Open it in your browser, select your agent from the menu, and start chatting. By default, this agent is configured to handle weather requests.<\/p>\n<figure class=\"\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/13d\/1d7\/0c6\/13d1d70c6d8168a1f3e181498e380f5c.png\" alt=\"The Playground\" sizes=\"(max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/\/post_images\/13d\/1d7\/0c6\/13d1d70c6d8168a1f3e181498e380f5c.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/13d\/1d7\/0c6\/13d1d70c6d8168a1f3e181498e380f5c.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/p>\n<div><figcaption>The Playground<\/figcaption><\/div>\n<\/figure>\n<p>When you are ready to evaluate, call:<\/p>\n<pre><code class=\"bash\">agents-cli eval run<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<h4>Deploy AI Agents<\/h4>\n<p>When you run\u00a0<code>agent-cli deploy<\/code>, you usually have two choices:<\/p>\n<ol>\n<li>\n<p><strong>Cloud Run:<\/strong>\u00a0This treats your agent like a standard web application. It\u2019s great if you want total control over the server.<\/p>\n<\/li>\n<li>\n<p><strong>Agent Runtime:<\/strong>\u00a0This is a specialized environment built specifically for AI. It handles things like Memory (remembering what the user said earlier) and Orchestration (managing multiple agents working together) automatically.<\/p>\n<\/li>\n<\/ol>\n<h4>The Deployment Process<\/h4>\n<ol>\n<li>\n<p>Ensure you have a\u00a0<strong>Google Cloud Project<\/strong>\u00a0set up.<\/p>\n<\/li>\n<li>\n<p>Enable billing in the Cloud Console.<\/p>\n<\/li>\n<li>\n<p>Run the command:<\/p>\n<\/li>\n<\/ol>\n<pre><code class=\"bash\"># Deploy to the configured target (Agent Runtime, Cloud Run or GKE)  agents-cli deploy<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>If you want to register your agent with Gemini Enterprise, run this command:<\/p>\n<pre><code class=\"bash\"># Register your deployed agent with Gemini Enterprise  agents-cli publish gemini-enterprise<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>It allows you to use your agent directly from\u00a0<a href=\"https:\/\/gemini.google.com\/\" rel=\"noopener noreferrer nofollow\">https:\/\/gemini.google.com\/<\/a><\/p>\n<p>Almost all commands accept flags such as\u00a0<code>--project<\/code>,\u00a0<code>--region<\/code>,\u00a0<code>--datastore<\/code>,\u00a0<code>--cicd-runner<\/code>\u00a0and\u00a0<code>--deployment-target<\/code>\u00a0so that you can customize the environment. Use\u00a0<code>agents-cli --help<\/code>\u00a0the\u00a0<a href=\"https:\/\/google.github.io\/agents-cli\/cli\/\" rel=\"noopener noreferrer nofollow\">CLI reference<\/a>\u00a0for full details.<\/p>\n<h3>Agents CLI Key Commands Cheat Sheet<\/h3>\n<ul>\n<li>\n<p><strong>agent-cli setup<\/strong>\u00a0Installs the core environment<\/p>\n<\/li>\n<li>\n<p><strong>agent-cli create [name]<\/strong>\u00a0Starts a new agent project<\/p>\n<\/li>\n<li>\n<p><strong>agent-cli playground<\/strong>\u00a0Launches the web testing interface<\/p>\n<\/li>\n<li>\n<p><strong>agent-cli eval run<\/strong>\u00a0Checks if your agent is performing correctly<\/p>\n<\/li>\n<li>\n<p><strong>agent-cli deploy<\/strong>\u00a0Pushes your agent to Google Cloud<\/p>\n<\/li>\n<\/ul>\n<h3>Building an agent with your coding assistant<\/h3>\n<p>After installation, open your coding assistant (Codex, Claude Code, Gemini CLI, or Copilot) and verify that the Agents CLI skills are visible:<\/p>\n<pre><code class=\"bash\">\/skills<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<figure class=\"\"><img decoding=\"async\" src=\"https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/532\/84a\/2db\/53284a2db66826e44c0e996d0a8bf149.png\" alt=\"Building an agent with your coding assistant\" sizes=\"(max-width: 780px) 100vw, 50vw\" srcset=\"https:\/\/habrastorage.org\/r\/w780\/getpro\/habr\/\/post_images\/532\/84a\/2db\/53284a2db66826e44c0e996d0a8bf149.png 780w,&#10;       https:\/\/habrastorage.org\/r\/w1560\/getpro\/habr\/\/post_images\/532\/84a\/2db\/53284a2db66826e44c0e996d0a8bf149.png 781w\" loading=\"lazy\" decode=\"async\"\/><\/p>\n<div><figcaption>Building an agent with your coding assistant<\/figcaption><\/div>\n<\/figure>\n<p>You should see\u00a0<code>google\u2011agents\u2011cli\u2011workflow<\/code>\u00a0and the other packaged skills. From there, you can instruct the agent using natural language. For example:<\/p>\n<pre><code class=\"bash\">Build a support agent that answers questions from our docs<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<p>Gemini CLI (or Claude Code, Codex, etc.) will call the CLI skills to scaffold the project, install dependencies, evaluate performance, and prepare deployment. This pattern works with any agent platform that supports skills installation.<\/p>\n<p>If you\u2019re happy with the result, simply call\u00a0<strong>deploy<\/strong>\u00a0command and your agent will be live:<\/p>\n<pre><code class=\"bash\"># Deploy to the configured target (Agent Runtime, Cloud Run or GKE)  agents-cli deploy<\/code><div class=\"code-explainer\"><a href=\"https:\/\/sourcecraft.dev\/\" class=\"tm-button code-explainer__link\" style=\"visibility: hidden;\"><img style=\"width:14px;height:14px;object-fit:cover;object-position:left;\"\/><\/a><\/div><\/pre>\n<h3>Video Tutorial: Agents CLI Explained<\/h3>\n<p>In this video, I\u2019ll show you how to use Google Agents CLI to build, test, and deploy AI agents from scratch.<\/p>\n<div class=\"tm-iframe_temp\" data-src=\"https:\/\/embedd.srv.habr.com\/iframe\/69f218824b66d902aa94d053\" data-style=\"\" id=\"69f218824b66d902aa94d053\" width=\"\" data-habr-games=\"\"><\/div>\n<p><strong><em>Watch on YouTube:<\/em><\/strong>\u00a0<a href=\"https:\/\/youtu.be\/C-0DIcFVt4Q\" rel=\"noopener noreferrer nofollow\"><strong><em>Agents CLI Explained<\/em><\/strong><\/a><\/p>\n<h3>Conclusion<\/h3>\n<p>Google\u2019s new agent\u2011oriented CLIs demonstrate how the command line is becoming a universal interface for both human operators and AI agents. Agents CLI offers a comprehensive, officially supported pathway to turn ideas into production\u2011ready agents on Google Cloud.<\/p>\n<p>Give it a shot and share with me what you build!<\/p>\n<p>Cheers, proflead! \ud83d\ude09<\/p>\n<\/div>\n<p>\u0441\u0441\u044b\u043b\u043a\u0430 \u043d\u0430 \u043e\u0440\u0438\u0433\u0438\u043d\u0430\u043b \u0441\u0442\u0430\u0442\u044c\u0438 <a href=\"https:\/\/habr.com\/ru\/articles\/1029726\/\">https:\/\/habr.com\/ru\/articles\/1029726\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Google Agents CLI Explained in 5 MinutesGoogle recently released the\u00a0Agentic CLI, a powerful tool that makes building, testing, and deploying AI agents faster and more intuitive. I think it\u2019s super useful, so I\u2019ll walk you through the entire lifecycle of an AI agent using the CLI.What is the Google Agents CLI?Before we dive into the code, let\u2019s understand the \u201cwhy.\u201d Usually, building an AI agent requires a lot of \u201cglue code\u201d, the messy scripts that connect an LLM (Large Language Model) to external tools or databases.What is the Google Agents CLIThe Agents CLI removes this friction. Agents CLI is an official Google tool for creating, evaluating, and deploying agents built with Google\u2019s Agent Development Kit (ADK). It acts as the programmatic backbone of the Agent Development Lifecycle (ADLC) on Google Cloud.Skills packaged for agentsAgents CLI bundles seven \u201cskills\u201d that teach coding agents how to perform each step of the ADLC:Skills packaged for agentsYour coding agent can call these skills directly once installed. Alternatively, a human developer can run the same commands manually (see \u201cHuman Mode\u201d below).How to Install Google Agents CLI?To get started, you need a clean environment. The CLI relies on a few modern web and data tools. Follow these steps to ensure everything is ready:PrerequisitesPython &amp; PIP:\u00a0The backbone of most AI development.sudo apt install python3 python3-pipNode.js &amp; NPM:\u00a0Required for the CLI\u2019s interface and deployment features.sudo apt install nodejs npmUV:\u00a0A lightning-fast Python package installer that the Agents CLI uses under the hood.pip install uvInstallation StepsCreate a Virtual Environment: It is always a good idea to keep your projects separate so they don\u2019t interfere with each other.python -m venv agent-env  source agent-env\/bin\/activate  # On Windows use `agent-env\\Scripts\\activate`The CLI is distributed through\u00a0uv:uvx google-agents-cli setupThe CLI supports macOS and Linux; native Windows is currently unsupported (use WSL 2 instead).AuthenticationAgents CLI picks up your credentials automatically if you are already authenticated with the\u00a0gcloud\u00a0CLI. When running locally without gcloud, you can set a\u00a0Gemini API key:export GEMINI_API_KEY=&#187;your\u2011key&#187;For detailed authentication scenarios (service accounts, A2A roles etc.) refer to the\u00a0Authentication guide. The important point is that your coding agent inherits whatever credentials your shell has; you do not need to embed secrets in code.Building Your First \u201cWeather\u201d AgentThe fastest way to learn is by doing. The CLI comes with a \u201cboilerplate\u201d or template system to get you moving in seconds.Run the following command in your terminal:agent-cli create my-first-agent  cd my-first-agentInstall DependenciesInside your new project folder, you need to install the specific libraries required for this agent:agent-cli installThe PlaygroundOne of the best features of this CLI is the Playground. Instead of testing your agent in a black-and-white terminal, you can launch a local web interface.agent-cli playgroundOnce you run this, you\u2019ll get a local URL (usually\u00a0http:\/\/127.0.0.1:8000). Open it in your browser, select your agent from the menu, and start chatting. By default, this agent is configured to handle weather requests.The PlaygroundWhen you are ready to evaluate, call:agents-cli eval runDeploy AI AgentsWhen you run\u00a0agent-cli deploy, you usually have two choices:Cloud Run:\u00a0This treats your agent like a standard web application. It\u2019s great if you want total control over the server.Agent Runtime:\u00a0This is a specialized environment built specifically for AI. It handles things like Memory (remembering what the user said earlier) and Orchestration (managing multiple agents working together) automatically.The Deployment ProcessEnsure you have a\u00a0Google Cloud Project\u00a0set up.Enable billing in the Cloud Console.Run the command:# Deploy to the configured target (Agent Runtime, Cloud Run or GKE)  agents-cli deployIf you want to register your agent with Gemini Enterprise, run this command:# Register your deployed agent with Gemini Enterprise  agents-cli publish gemini-enterpriseIt allows you to use your agent directly from\u00a0https:\/\/gemini.google.com\/Almost all commands accept flags such as\u00a0&#8212;project,\u00a0&#8212;region,\u00a0&#8212;datastore,\u00a0&#8212;cicd-runner\u00a0and\u00a0&#8212;deployment-target\u00a0so that you can customize the environment. Use\u00a0agents-cli &#8212;help\u00a0the\u00a0CLI reference\u00a0for full details.Agents CLI Key Commands Cheat Sheetagent-cli setup\u00a0Installs the core environmentagent-cli create [name]\u00a0Starts a new agent projectagent-cli playground\u00a0Launches the web testing interfaceagent-cli eval run\u00a0Checks if your agent is performing correctlyagent-cli deploy\u00a0Pushes your agent to Google CloudBuilding an agent with your coding assistantAfter installation, open your coding assistant (Codex, Claude Code, Gemini CLI, or Copilot) and verify that the Agents CLI skills are visible:\/skillsBuilding an agent with your coding assistantYou should see\u00a0google\u2011agents\u2011cli\u2011workflow\u00a0and the other packaged skills. From there, you can instruct the agent using natural language. For example:Build a support agent that answers questions from our docsGemini CLI (or Claude Code, Codex, etc.) will call the CLI skills to scaffold the project, install dependencies, evaluate performance, and prepare deployment. This pattern works with any agent platform that supports skills installation.If you\u2019re happy with the result, simply call\u00a0deploy\u00a0command and your agent will be live:# Deploy to the configured target (Agent Runtime, Cloud Run or GKE)  agents-cli deployVideo Tutorial: Agents CLI ExplainedIn this video, I\u2019ll show you how to use Google Agents CLI to build, test, and deploy AI agents from scratch.Watch on YouTube:\u00a0Agents CLI ExplainedConclusionGoogle\u2019s new agent\u2011oriented CLIs demonstrate how the command line is becoming a universal interface for both human operators and AI agents. Agents CLI offers a comprehensive, officially supported pathway to turn ideas into production\u2011ready agents on Google Cloud.Give it a shot and share with me what you build!Cheers, proflead! ;)\u0441\u0441\u044b\u043b\u043a\u0430 \u043d\u0430 \u043e\u0440\u0438\u0433\u0438\u043d\u0430\u043b \u0441\u0442\u0430\u0442\u044c\u0438 https:\/\/habr.com\/ru\/articles\/1029726\/<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[],"tags":[],"class_list":["post-477983","post","type-post","status-publish","format-standard","hentry"],"_links":{"self":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts\/477983","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=477983"}],"version-history":[{"count":0,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=\/wp\/v2\/posts\/477983\/revisions"}],"wp:attachment":[{"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=477983"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=477983"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/savepearlharbor.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=477983"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}