Google Skills Repository Categories: Complete Guide to 12 Skill Groups

The google/skills repository organizes its skills into 12 high-level categories defined in the README.md, ranging from "Getting started with Google Cloud" to "Advertising" and "Others," with each skill's metadata.category field determining its group membership.

The google/skills repository is Google's open-source collection of reusable skill definitions for AI agents and automation workflows. Understanding the categories of skills in the google/skills repository helps developers quickly locate relevant capabilities for their Google Cloud projects.

---

How Skill Categories Are Defined

Categories in the google/skills repository originate from two sources:

1. Metadata fields in individual skill definition files
2. README.md section headers that group skills visually

Each skill is defined in a SKILL.md file containing YAML front-matter with a metadata.category field. The repository's root README.md then aggregates these into readable groups.

---

Complete List of Google Skills Repository Categories

Getting Started with Google Cloud

Introductory skills covering authentication, onboarding workflows, and foundational Google Cloud concepts. These skills target new users establishing their first GCP environment.

Located in README.md lines 23-27.

Multi-Product Solution Skills

End-to-end solutions that orchestrate multiple Google Cloud products into cohesive workflows. These skills demonstrate cross-product integration patterns rather than single-service capabilities.

Located in README.md lines 27-35.

AI/ML

Skills exposing Gemini, Agent Platform, and other artificial intelligence and machine learning APIs. This rapidly expanding category includes model invocation, prompt engineering, and agent orchestration primitives.

Located in README.md lines 37-53.

Infrastructure

Core infrastructure skills spanning:

• GKE (Google Kubernetes Engine)
• Terraform configuration management
• Networking and connectivity

Located in README.md lines 56-78.

Databases and Analytics

Data platform skills including BigQuery, Spanner, Bigtable, AlloyDB, and associated analytics tooling. These skills handle data ingestion, querying, and database administration tasks.

Located in README.md lines 85-96.

Developer Tools

Utilities for developer productivity:

• gcloud CLI operations
• Google Agents CLI interactions

Located in README.md lines 97-99.

Management Tools

Operational management capabilities covering monitoring, logging, cost analysis, and other observability functions. These skills help maintain production health and optimize resource spending.

Located in README.md lines 101-114.

Well-Architected Framework

Skills explicitly aligned with Google's WAF pillars:

• Cost optimization
• Operational excellence
• Performance efficiency
• Reliability
• Security
• Sustainability

Located in README.md lines 115-121.

Security and Identity

Platform-level security, workload security, and detection-coverage skills. These include identity management, access control, and threat detection capabilities.

Located in README.md lines 122-126.

Web and App Hosting

Deployment and hosting skills for Cloud Run and Firebase platforms. Simplifies containerized and serverless application delivery.

Located in README.md lines 127-128.

Advertising

Integration skills for Google's advertising stack:

• Google Mobile Ads SDK
• IMA (Interactive Media Ads) SDK

Located in README.md lines 130-143.

Others

Miscellaneous skills not fitting above categories, including:

• Google Analytics Admin/Data APIs
• External links to other Google-maintained skill collections

Located in README.md lines 144-146.

---

Programmatically Extracting Skill Categories

You can dynamically discover categories of skills in the google/skills repository by parsing the YAML front-matter from SKILL.md files:

import yaml
import os
import glob

def list_skill_categories(repo_root: str) -> set:
    """
    Extract unique categories from all SKILL.md files in the repository.
    
    Each skill definition contains metadata.category in its YAML front-matter.
    """
    categories = set()
    
    for path in glob.glob(
        os.path.join(repo_root, '**/SKILL.md'), 
        recursive=True
    ):
        with open(path) as f:
            # Extract YAML front-matter (delimited by ---)

            lines = []
            for line in f:
                if line.strip() == '---' and lines:
                    break
                lines.append(line)
            
            front = yaml.safe_load('\n'.join(lines))
            if front and 'metadata' in front:
                category = front['metadata'].get('category')
                if category:
                    categories.add(category)
    
    return categories


# Example usage against local clone

repo_root = "/path/to/google/skills"
print(sorted(list_skill_categories(repo_root)))

This matches the 12 categories enumerated in the repository's README.md, confirming the categorization is data-driven from individual skill metadata.

---

Key Files Defining the Categorization Structure

File Path	Purpose
README.md	Central index grouping all skills by category with descriptive headers
skills/**/SKILL.md	Individual skill definitions containing metadata.category fields
skills/**/references/*	Supporting reference materials linked to categorized skills

The relationship between metadata.category values and README.md headings ensures consistent organization across the repository.

---

Summary

• 12 defined categories span from introductory cloud skills to specialized advertising integrations
• Category assignment flow: SKILL.md metadata → README.md grouping headers
• Source of truth: metadata.category field in each skill's YAML front-matter
• Primary documentation: README.md at repository root (lines 23-146)
• Programmatic access: Parse YAML front-matter from **/SKILL.md files

---

Frequently Asked Questions

How do I find which category a specific skill belongs to?

Check the metadata.category field in the skill's SKILL.md file. This field determines where the skill appears in the README.md grouping. You can also search the README for the skill name to locate its assigned category header.

Can skills belong to multiple categories?

No. The current implementation in the google/skills repository uses a single metadata.category string per skill. Skills requiring cross-category description are typically placed under Multi-product solution skills or listed under their primary functional area.

What is the relationship between Well-Architected Framework skills and other categories?

Well-Architected Framework skills are cross-cutting concerns that may overlap with infrastructure, security, or management tools, but are explicitly tagged to align with Google's WAF pillars. A skill optimizing BigQuery costs, for example, could appear under both Databases and analytics and Well-Architected Framework.

How often do the categories change?

Category definitions are version-controlled in README.md and evolve with Google Cloud's product portfolio. Major additions like the AI/ML expansion (lines 37-53) reflect new platform priorities. Monitor the repository's commit history to README.md for structural changes.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}