What is google/skills?

google/skills is a repository of agent skill definitions that teach AI coding assistants how to work with Google Cloud products and APIs. Engineers add specific skills to their AI coding environment with a single npx command, and the assistant gains structured guidance for tasks like GKE cluster creation, Gemini API usage, or Cloud Storage configuration.

Can I use google/skills commercially?

Yes. Apache-2.0 is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.

Is google/skills still maintained?

Yes. The repository last received commits 4 days ago.

What programming language is google/skills written in?

Mainly Python, according to GitHub's language statistics.

google/skills: Google Cloud Agent Skills for Claude Code

At a glance

What is it?: google/skills is a repository of agent skill definitions that teach AI coding assistants how to work with Google Cloud products and APIs. Engineers add specific skills to their AI coding environment with a single npx command, and the assistant gains structured guidance for tasks like GKE cluster creation, Gemini API usage, or Cloud Storage configuration.
Who is it for?: Engineers who work with Google Cloud and use an AI coding assistant that supports the Agent Skills protocol will find this repository a direct productivity gain for tasks like GKE cluster setup, Agent Platform deployment, or Genkit development. Teams that do not use Google Cloud, or whose AI coding assistant does not support the npx skills install flow, will get nothing from the repository.
Can I use it commercially?: Yes. Apache-2.0 is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
Is it still maintained?: Yes. The repository last received commits 4 days ago.
What is it written in?: Mainly Python, according to GitHub's language statistics.

Answers come from the project's GitHub data, last synced on September 25, 2026, and from our analysis. They are not legal advice.

Editorial analysis

Written by the Hysen Labs editorial pipeline with AI assistance. Checked against the repository by machine, not yet reviewed by a named editor. How we write these Last updated August 4, 2026.

What Agent Skills Are and Why This Repository Exists

AI coding assistants can follow instructions embedded in skill definition files to produce more accurate, task-specific guidance. google/skills provides those definition files for Google Cloud and related Google products. The README describes the repository as providing 'agent instructions for working with Google Cloud and other Google products through supported tools and APIs.' The skills are not code libraries or SDK wrappers. They are structured instruction sets that tell an AI assistant what steps to take, which APIs to call, and which common mistakes to avoid when helping with a specific Google Cloud task. Without skills, an assistant relies on its general training data, which may lag behind current API versions or miss Google Cloud-specific conventions. With skills installed, the assistant has explicit, maintained guidance it can follow.

How the Repository Is Organized

The top-level directory contains skills/, which holds all the individual skill definitions organized by category and product. Each skill lives in its own subdirectory under skills/cloud/ or skills/developers/. The repository also contains .claude-plugin/ and .agents/ directories, which are the integration points for specific AI coding assistants. The index.json file at the top level acts as a registry listing all available skills. Categories span Getting Started with Google Cloud (auth, onboarding, foundation builder), Multi-product solutions (solution architecture, multi-agent security, analytics, data lakehouse, RAG), AI and ML (Agent Platform alert configuration, endpoint management, eval flywheel, model tuning, Gemini API, Genkit in Dart, Go, JS, and Python, LiveAPI), and Infrastructure (GKE cluster creation, GKE networking, GKE storage, GKE upgrades, Cloud Storage, Filestore, load balancing). There is also a Google Skill Finder skill under skills/developers/ that helps users discover which skills apply to their task.

Installing Google Skills in Your AI Coding Assistant

The README gives a single install command:

bash

npx skills add google/skills

Running this command presents a selection prompt where you choose which specific skills from the repository to install. You do not have to install all of them at once. The command is handled by the skills CLI (available at skills.sh/google/skills per the README badge), which downloads the selected skill definitions and places them where your AI coding assistant can find them. After installation, the assistant can load the relevant skill when you ask it to perform a task covered by that skill. The README does not document what happens if your assistant does not support the skills format, so verify compatibility before running the install.

What Each Category Covers in Practice

The GKE skills are the largest group and the most specific. They cover discrete operational tasks: creating a cluster, configuring autoscaling, setting up networking, managing storage, troubleshooting node readiness failures, troubleshooting workload scaling problems, and handling batch and HPC workloads. Each skill is a separate unit, so installing only gke-cluster-creation does not pull in gke-networking or gke-storage. The Agent Platform skills cover the Gemini Enterprise Agent Platform (formerly Vertex AI): deploying models, managing endpoints, configuring the RAG engine, running evaluations, and migrating from AI Studio. The Genkit skills cover developing with Genkit in four languages: Dart, Go, JavaScript, and Python. The multi-product solution skills describe architecture workflows that combine several Google Cloud products, such as building a hybrid search system with AlloyDB or deploying a secure n-tier serverless application.

Limitations: What the Skills Do Not Cover

The repository covers Google Cloud and a subset of other Google developer products. It does not include skills for third-party cloud providers, non-Google APIs, or general software engineering tasks. The skills are instruction sets, not executable code, so they cannot run commands on your behalf or authenticate you to Google Cloud. Authentication is a prerequisite, not something a skill handles. The README lists a dedicated skill for Google Cloud authentication (google-cloud-recipe-auth), but installing that skill instructs the assistant on how to help you authenticate rather than performing authentication itself. Skills also cannot guarantee that the AI assistant will follow them correctly every time; the output quality depends on the assistant's ability to interpret and apply the skill definitions. The repository has no GitHub releases, so there is no version history to pin against if a skill's content changes.

Maintenance and Licence

The last push to the repository was on 2026-09-25, which is three days before this article was written. The repository is licensed under Apache 2.0, permitting use, modification, and redistribution with attribution. Because skills are instruction text rather than compiled code, updates are typically additive: new skills appear as new subdirectories, and existing skills change when Google Cloud APIs or best practices change. Contributors follow the contributing guidelines in CONTRIBUTING.md. The .gitmodules file indicates the repository uses Git submodules, which means some skill content may live in separate repositories that are included here as submodules. Cloning with submodules requires the --recurse-submodules flag to get full content.

Editorial conclusion

Engineers who work with Google Cloud and use an AI coding assistant that supports the Agent Skills protocol will find this repository a direct productivity gain for tasks like GKE cluster setup, Agent Platform deployment, or Genkit development. Teams that do not use Google Cloud, or whose AI coding assistant does not support the npx skills install flow, will get nothing from the repository. Before installing, confirm that your assistant supports the .claude-plugin/ or .agents/ format used here, since the README does not list every compatible tool explicitly.

Frequently asked questions

How do I install skills in Claude Code?

Run `npx skills add google/skills` in your terminal. The command presents a selection prompt where you choose which specific Google Cloud skills to install. After installation, your Claude Code session can use those skills when you ask it to perform covered tasks.

How do I use skills in Claude Code?

After installing with `npx skills add google/skills`, ask your Claude Code assistant to perform a task covered by an installed skill, such as creating a GKE cluster or configuring a Gemini API endpoint. The assistant loads the relevant skill definition and follows its guidance.

How do I install skills from GitHub to Claude?

The README gives the command `npx skills add google/skills`, which handles downloading and installing the skill definitions from the GitHub repository into your AI coding assistant environment. The skills CLI manages the process of selecting and placing the files.

google/skills: Agent Instructions for Google Cloud in AI Coding Assistants