Where to Find Google Cloud Product Skill Definitions in the google/skills Repository

Google Cloud product skill definitions are located in SKILL.md files within product-specific subdirectories under skills/cloud/ in the google/skills repository.

The google/skills repository provides structured automation definitions for Google Cloud workflows. Each Cloud product or feature has a dedicated skill package containing a SKILL.md file that specifies inputs, outputs, and reference materials for programmatic interaction.

Location and Directory Structure

Root Path for Cloud Skills

All Google Cloud skill definitions reside in the skills/cloud/ directory. This path serves as the root for every Cloud-related skill module in the repository.

Standard Package Layout

Each product is organized as a self-contained skill package under skills/cloud/<product-name>/. According to the repository structure, a typical skill directory contains:

• SKILL.md — The authoritative definition file describing the skill's purpose, parameters, and usage
• references/ — Optional directory containing supplementary documentation, CLI usage guides, and IAM role references
• assets/ — Optional directory containing YAML policy templates, deployment manifests, and configuration scripts

Examples of Google Cloud Product Skills

The repository contains skill definitions for major Google Cloud products. Each product has a dedicated subdirectory under skills/cloud/ containing its SKILL.md file:

Product	Directory Path	Definition File
Cloud Run Basics	skills/cloud/cloud-run-basics/	SKILL.md
BigQuery AI & ML	skills/cloud/bigquery-ai-ml/	SKILL.md
GKE Compute Classes	skills/cloud/gke-compute-classes/	SKILL.md
GKE Workload Security	skills/cloud/gke-workload-security/	SKILL.md
IAM Helper for Privileged Access Management	skills/cloud/iam-helper-for-privileged-access-management/	SKILL.md

Anatomy of a SKILL.md File

Every SKILL.md follows a standardized structure to ensure consistent programmatic parsing. The file contains:

• Description — Top-level overview of the skill's purpose and capabilities
• Inputs — Parameter definitions required to execute the skill
• Outputs — Expected results or return values
• References — Pointers to additional documentation, example scripts, or external resources

Accessing Skill Definitions Programmatically

You can load and utilize these skill definitions directly from the repository structure.

Loading a Skill Definition in Python

To read a skill definition programmatically, navigate to the skills/cloud/ hierarchy and load the SKILL.md file:

import pathlib

def load_skill(product):
    """Return the raw Markdown of a Cloud skill."""
    base = pathlib.Path(__file__).parent.parent / "skills" / "cloud" / product / "SKILL.md"
    return base.read_text(encoding="utf-8")

print(load_skill("cloud-run-basics"))

Using Skills in GitHub Actions Workflows

Reference skill definitions in automation workflows by specifying the path to the SKILL.md file:

- name: Deploy Cloud Run service
  uses: google/skills@main
  with:
    skill_path: skills/cloud/cloud-run-basics/SKILL.md
    # Pass required inputs defined in the SKILL.md

    service_name: my-service
    image: gcr.io/my-project/my-image:latest

Accessing Skill Assets from Scripts

Apply configuration assets shipped with a skill by referencing the assets/ subdirectory:

# Example: apply a GKE policy asset shipped with a skill

kubectl apply -f \
  "$(git rev-parse --show-toplevel)/skills/cloud/gke-workload-security/assets/default-deny-netpol.yaml"

Summary

• Google Cloud skill definitions are stored as SKILL.md files in the skills/cloud/ directory of the google/skills repository.
• Each product has a dedicated subdirectory (e.g., skills/cloud/cloud-run-basics/) containing the definition and optional supporting files.
• The SKILL.md file structure includes standardized sections for inputs, outputs, and references.
• Supplementary materials reside in references/ (documentation) and assets/ (configuration files) subdirectories.
• You can access these definitions programmatically using standard file I/O or through workflow automation tools.

Frequently Asked Questions

What is the exact file path pattern for Google Cloud skill definitions in the repository?

Google Cloud product skill definitions follow the pattern skills/cloud/<product-name>/SKILL.md. For example, the Cloud Run Basics skill definition is located at skills/cloud/cloud-run-basics/SKILL.md. The skills/cloud/ directory serves as the root container for all Cloud-related skill packages.

What sections are included in a SKILL.md file?

Every SKILL.md file contains a top-level description, a list of inputs (parameters), outputs, and a references section. The references section points to additional documentation, example scripts, or YAML assets stored in the accompanying references/ and assets/ directories.

Can I use skill definitions in automated workflows?

Yes. As implemented in google/skills, you can reference skill definitions in automation workflows by specifying the skill_path parameter (e.g., skills/cloud/cloud-run-basics/SKILL.md) and passing required inputs defined in the file. The repository structure supports direct integration with GitHub Actions and other CI/CD platforms.

Where are supporting assets like YAML templates and scripts stored?

Supporting assets are stored in references/ and assets/ subdirectories within each product folder. For example, GKE Workload Security includes policy templates in skills/cloud/gke-workload-security/assets/, while IAM Helper stores documentation in its references/ folder.

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