What Product Categories Are Supported by Google Skills? Complete 2024 Guide

Google Skills organizes its automation components into 15 distinct product categories spanning the entire Google Cloud portfolio, from core infrastructure to AI/ML and security domains.

The google/skills repository structures reusable conversational automation components—called "skills"—into logical product categories defined within individual SKILL.md front-matter. Each category groups related capabilities that align with specific Google Cloud service domains, making it easier to discover and invoke relevant automation for your use case.

How Product Categories Are Defined in Google Skills

According to the source code structure, every skill declares its product category within the YAML front-matter of its SKILL.md file. The repository uses a standardized category: key to classify each skill.

For example, a skill targeting observability features includes:

category: CloudObservabilityAndMonitoring

This front-matter declaration enables automated categorization across the repository. The grep output confirms this pattern appears consistently across all skill definitions, allowing both human users and automation tools to filter capabilities by domain.

Complete List of Supported Product Categories

The repository currently supports the following 15 product categories, each representing a major Google Cloud domain:

• CloudInfrastructureAndServices: Core infrastructure services including Compute Engine, Cloud Run, and gcloud CLI tooling. Example: skills/cloud/gcloud/SKILL.md

• MultiProductSolutions: End-to-end solutions integrating multiple Google Cloud products for complex scenarios. Example: skills/cloud/google-cloud-solution-multi-agent-security/SKILL.md

• CloudObservabilityAndMonitoring: Logging, tracing, alerting, and workload observability. Example: skills/cloud/gke-observability/SKILL.md

• Databases: Managed database services and operational best practices. Example: skills/cloud/spanner-basics/SKILL.md

• BigDataAndAnalytics: Data processing pipelines and analytics platforms. Example: skills/cloud/managed-airflow-migrations/SKILL.md

• Security: IAM management, policy simulation, threat detection, and security posture. Example: skills/cloud/iam-helper-for-policy-simulator/SKILL.md

• WellArchitectedFramework: Guidance aligned with Google Cloud's Well-Architected Framework pillars—security, reliability, cost optimization, performance, sustainability, and operational excellence. Example: skills/cloud/google-cloud-waf-security/SKILL.md

• Storage: Cloud Storage management, bucket lifecycle policies, and performance optimization. Example: skills/cloud/google-cloud-storage-basics/SKILL.md

• Compute: Compute-oriented services including Compute Engine, GPUs, and TPUs. Example: skills/cloud/google-cloud-networking-observability/SKILL.md

• Networking: VPC configuration, Service Networking, load balancing, and hybrid connectivity. Example: skills/cloud/gke-service-networking/SKILL.md

• Containers: GKE, Anthos, and container lifecycle management. Example: skills/cloud/gke-basics/SKILL.md

• DevOps: CI/CD pipelines, Cloud Build integration, and developer tooling. Example: skills/cloud/google-agents-cli-onboarding/SKILL.md

• AiAndMachineLearning: Generative AI, Vertex AI, Gemini integration, and related SDKs. Example: skills/cloud/genkit-python/SKILL.md

• GettingStarted: Introductory recipes and onboarding flows for new Google Cloud users. Example: skills/cloud/google-cloud-recipe-onboarding/SKILL.md

• CloudInfrastructure: Infrastructure-specific automation and configuration tasks such as GKE alert configuration. Example: skills/cloud/gke-alert-configuration/SKILL.md

Invoking Skills by Product Category

While categories organize discovery, you invoke specific skills by their unique identifiers. The invocation pattern varies by agent framework but follows consistent semantics across the repository.

Use the Python SDK to call observability skills:

# Example: invoke the GKE observability skill

result = agent.run_skill(
    skill_id="gke-observability",
    query="Show me recent CPU usage for my GKE cluster"
)
print(result)

Execute AI/ML skills via CLI:

# Example: invoke the Genkit Python skill via the CLI

genkit run --skill genkit-python --input "Generate a Python function that adds two numbers"

Integrate Well-Architected Framework checks into YAML workflows:

# Example: invoke a Well-Architected Framework skill in a YAML-based workflow

steps:
  - name: check-security-posture
    uses: google/cloud-waf-security@v1
    with:
      workload: my-app

Key Source Files by Category

The following SKILL.md files serve as canonical entry points for their respective product categories:

• CloudInfrastructureAndServices: skills/cloud/gcloud/SKILL.md
• MultiProductSolutions: skills/cloud/google-cloud-solution-multi-agent-security/SKILL.md
• CloudObservabilityAndMonitoring: skills/cloud/gke-observability/SKILL.md
• Databases: skills/cloud/spanner-basics/SKILL.md
• BigDataAndAnalytics: skills/cloud/managed-airflow-migrations/SKILL.md
• Security: skills/cloud/iam-helper-for-policy-simulator/SKILL.md
• WellArchitectedFramework: skills/cloud/google-cloud-waf-security/SKILL.md
• Storage: skills/cloud/google-cloud-storage-basics/SKILL.md
• Compute: skills/cloud/google-cloud-networking-observability/SKILL.md
• Networking: skills/cloud/gke-service-networking/SKILL.md
• Containers: skills/cloud/gke-basics/SKILL.md
• DevOps: skills/cloud/google-agents-cli-onboarding/SKILL.md
• AiAndMachineLearning: skills/cloud/genkit-python/SKILL.md
• GettingStarted: skills/cloud/google-cloud-recipe-onboarding/SKILL.md
• CloudInfrastructure: skills/cloud/gke-alert-configuration/SKILL.md

Summary

• Google Skills supports 15 product categories covering the full Google Cloud portfolio.
• Categories are declared via the category: field in each skill's SKILL.md front-matter.
• Domains range from CloudInfrastructureAndServices and Compute to AiAndMachineLearning and WellArchitectedFramework.
• Skills are invoked by specific ID rather than category, using Python, CLI, or YAML workflow patterns.
• Each category contains canonical example skills with documented entry points in the repository.

Frequently Asked Questions

How do I determine which product category a skill belongs to?

Check the YAML front-matter in the skill's SKILL.md file. The category: field explicitly declares the product category, as implemented in the google/skills repository structure. You can also search the repository for category: entries to discover skills by domain.

Can a single skill belong to multiple product categories?

Based on the source code structure, each SKILL.md file specifies a single category: value in its front-matter. Skills are organized into one primary category to maintain clear organizational boundaries, though MultiProductSolutions skills may reference capabilities across domain boundaries.

What is the difference between CloudInfrastructure and CloudInfrastructureAndServices?

CloudInfrastructureAndServices covers broad infrastructure services including Compute Engine, Cloud Run, and gcloud CLI tooling. CloudInfrastructure focuses on specific infrastructure configuration tasks such as GKE alert configuration and detailed infrastructure automation, representing a more specialized subset of infrastructure management.

How do I invoke skills programmatically from a specific category?

Use the skill's unique identifier rather than the category name. For Python applications, use agent.run_skill() with the specific skill_id. For CLI workflows, use framework-specific commands like genkit run --skill [id]. For CI/CD pipelines, reference the skill in YAML workflow steps using the uses: directive with the appropriate path or action reference.

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