Guide · Agent skill

How to use the Skill Registry (Google Cloud) agent skill

Updated 2026-10-29 · written with AI from the project's own documentation and checked against our catalogue data

The Skill Registry (Google Cloud) agent skill teaches an AI agent how to work with Google Cloud's Skill Registry, a hosted catalog where agent skills are stored and searched. This guide is for developers and platform teams who already have, or can set up, a Google Cloud project with the Agent Platform enabled. By the end you will know what the skill does, how to install it, how to ask for things in plain language, and what safety steps to take before changing anything in a shared registry.

Skill Registry (Google Cloud)google/skills · Apache-2.0 Download the skill

Quick start

1Install the prerequisites. Install Python 3 and the pip packages google-auth and requests, and install the gcloud CLI.
2Authenticate with Google Cloud. Run gcloud auth application-default login so the scripts can use your credentials.
3Set your environment variables. Set GCP_PROJECT_ID to your project ID and GCP_LOCATION to a region such as us-central1.
4Install the skill. Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
5Validate the environment. Run python3 scripts/validate_env.py and fix any reported problems.
6Ask for a registry task. Describe the search, listing or upload you need in plain language, and review the agent's command before it runs.

What is an agent skill, and what does this one do?

An agent skill is a folder of instructions and optional scripts that an AI agent loads when a task matches it. The main file is SKILL.md, which tells the agent what the skill is for and how to use it.

Once this skill is installed, the agent knows how to talk to the Skill Registry on the Gemini Enterprise Agent Platform. It can search skills by keyword, list them, get a specific skill, inspect revision histories, upload new skills, update or permanently delete existing ones, check on long-running operations, and scaffold a new skill folder on your local machine.

Skill discovery: search, list, get and inspect revision history
Skill lifecycle: upload, update or permanently delete
Operation monitoring: check whether long-running changes have completed
Generate skill: create a starter skill folder locally

What you need before you install

The skill runs bundled Python scripts that call Google Cloud APIs using your own credentials, so nothing works until the environment is ready. The skill's own requirements list these items.

You need Python 3, the pip packages google-auth and requests (also listed in scripts/requirements.txt), the gcloud CLI, and a Google Cloud project with Gemini Enterprise Agent Platform access. You also need to set the environment variables GCP_PROJECT_ID (your project ID) and GCP_LOCATION (a region such as us-central1).

Python 3
pip packages: google-auth, requests
gcloud CLI with application-default login
Google Cloud project with Gemini Enterprise Agent Platform access
Environment variables: GCP_PROJECT_ID and GCP_LOCATION

pip install google-auth requests
gcloud auth application-default login

How do you install the skill?

Download the skill's ZIP from this page and unzip it. Put the folder in your personal skills directory, ~/.claude/skills/, so it is available in every project, or in a project's .claude/skills/ folder so it applies only to that project.

If you use the Claude apps instead, upload the ZIP under Settings → Capabilities → Skills.

Personal install: unzip into ~/.claude/skills/
Project install: unzip into the project's .claude/skills/ folder
Claude apps: Settings → Capabilities → Skills, then upload the ZIP

How do you check that the environment is ready?

The skill's core directive is that the agent must run the environment validation check before any operation. You can also run it yourself from the skill folder to confirm your setup before asking for work.

Run the validation script first. If it reports a problem, fix it before searching or changing anything in the registry.

python3 scripts/validate_env.py

How do you trigger the skill with a natural request?

You do not need to type a command name. Describe what you want in plain language that relates to Skill Registry work, such as searching for a capability, listing skills, or checking an operation. The agent will load the skill when the request matches its description.

For the first search, the skill's quickstart uses a keyword query with a limit on the number of results. A request like the one below is a good starting point.

Search the Skill Registry for skills about test automation and show the top five.
List the skills in my Google Cloud registry and tell me which ones were updated recently.
Show the revision history for the skill with ID my-sample-skill.
Check whether my last upload operation has finished.

python3 scripts/skill_registry_ops.py search \
  --query "test skill" \
  --top-k 5

How do you upload, update or delete a skill?

Uploading takes either a folder or a ZIP file. The upload command requires a skill ID, a display name, a description, and exactly one of --folder or --zip-file. The source recommends the folder option. Be careful with deletion: the skill can permanently delete registry skills and upload folder contents, which can be irreversible.

Test on a non-production project first, and review what will be uploaded before confirming.

--skill-id (required): unique identifier for the skill
--display-name (required): human-readable name
--description (required): what the skill does
--folder or --zip-file (required, mutually exclusive): local folder or .zip file

python3 scripts/skill_registry_ops.py upload \
  --skill-id "my-sample-skill" \
  --display-name "My Sample Skill" \
  --description "A test skill uploaded

What safety steps should you take?

The skill runs scripts and uses the network. It calls Google Cloud APIs with your own credentials, so anything it does is done as you. Read the scripts before running them on a project you care about, and keep a human in the loop for any upload, update or delete.

Attribution tagging is part of the skill's instructions: terminal gcloud commands are prefixed with a metrics environment variable that identifies only the skill and its version. The source says not to use gcloud config set for this, because it would persist beyond the task.

Use a test project before any shared or production project
Review folder contents before uploading, since they may be exposed to the registry
Deletion is permanent; confirm the skill ID first
Do not store credentials in the skill folder; rely on gcloud application-default login

What people use it for

Searching a shared registry for existing skills before building a duplicate capability
Publishing a team's internal skill so other developers can discover it
Listing and checking the revision history of skills a platform team maintains
Monitoring whether a long-running upload or update has completed
Scaffolding a new skill folder locally as a starting draft

Frequently asked questions

How do I install the Skill Registry agent skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project. In the Claude apps, you can instead upload the ZIP under Settings → Capabilities → Skills.

What do I need to use the Skill Registry skill?

You need Python 3, the google-auth and requests packages, the gcloud CLI with application-default login, and a Google Cloud project with Gemini Enterprise Agent Platform access. You also need GCP_PROJECT_ID and GCP_LOCATION set.

Is the Skill Registry skill free?

The skill is published under the Apache-2.0 licence. The source does not state pricing for Google Cloud usage, so check Google Cloud's own billing terms for your project.

Can I use this without a Google Cloud account?

No. The scripts need GCP_PROJECT_ID, GCP_LOCATION and valid credentials, so nothing works without a Google Cloud project.

Why does the skill say to run validate_env.py first?

The skill requires an environment check before any operation. Running python3 scripts/validate_env.py catches missing variables or credentials before a search or upload fails partway through.

Can the skill delete things by accident?

It can permanently delete registry skills, which may be irreversible. Confirm the skill ID and use a test project before running any delete or upload on shared work.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits developers and platform teams on Google Cloud who want to publish, search and manage agent skills in a shared registry, and who are comfortable with gcloud. If you do not have a Google Cloud project, or you only want help writing or editing documents, pick a different tool.

Download the skill