Gcloud Cli

Wraps the gcloud CLI for Google Cloud operations with some sensible guardrails built in. Covers compute instances, storage buckets, IAM policies, and logging with progressive disclosure tables showing what needs confirmation (VM deletion) and what's outright blocked (project deletion, IAM policy removal). The memory protocol is notably strict about documenting learnings and decisions, which makes sense given how easy it is to wreck a GCP project if you're not careful. Main value is the 90%+ context savings claim and not having to remember gcloud's flag syntax. Best paired with a devops agent for cloud operations. Assumes you've already got gcloud installed and authenticated, so it's really just orchestration over an existing setup.

skills.sh installs: 999
GitHub stars: 33
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add oimiragieo/agent-studio --skill gcloud-cli --agent claude-code

Installs into .claude/skills of the current project.

Files

Google Cloud CLI Skill

Installation

The skill invokes the gcloud CLI. Install and initialize:

Linux/macOS: curl https://sdk.cloud.google.com | bash then restart shell and run gcloud init
Windows: Download Google Cloud SDK installer or use PowerShell to download and run

Verify: gcloud --version. Configure: gcloud init

Cheat Sheet & Best Practices

Config: gcloud config set project <id>; gcloud config list; gcloud config configurations create/activate <name> — switch projects.

Auth: gcloud auth login; use --impersonate-service-account for SA; --access-token-file for CI.

Hacks: Use --format="table(name,zone,status)" or --format=json + jq to cut output. Use --flags-file=file.yaml for long or repeated flags. Use named configurations for dev/staging/prod. Run gcloud components update periodically.