A collection of Agent Skills for Google products and technologies, including Google Cloud. This is my fork of google/skills, kept as a reference for how Google structures skills and as a source for the ones I install.
The concepts
A skill is a folder of instructions an AI coding agent loads only when a task calls for it. Instead of hoping the model remembers the right way to set up IAM or deploy to GKE, you give it a reviewed, versioned recipe. The collection is organized the way the work is:
- Getting started — authenticating to Google Cloud, onboarding, and a foundation builder.
- Multi-product solutions — a solution-architecture workflow, multi-agent security with Agent Gateway, building and deploying AI agents, hybrid search on AlloyDB, RAG on GKE, and a secure n-tier serverless web app.
- AI/ML — Agent Platform skills for inference, endpoint management, evaluation, alerting, and Model Garden deployment.
There is also a “skill finder” skill, which helps an agent locate the right skill for a task.
How to use it
npx skills add google/skills
The installer asks which skills you want, which agents to configure, and whether to install for the project or globally. To update later, and optionally remove obsolete skills:
npx skills update
A skill is instructions your agent will follow with your credentials, so install from the upstream repository and read a skill before relying on it.
Language: Python
Source code: github.com/avnit/skills
Upstream: github.com/google/skills