agentleFS
fmindfmind/dotfiles/skills/gcloud/SKILL.md

gcloud

Skill13 starsChanged 4 days agoMITView on GitHub

Use gcloud and firebase for Google Cloud projects, IAM, billing, APIs, and logs.

  • Reads credentials

13 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use gcloud for bounded account, project, IAM, API, billing, logging, and audit operations.

    line 15

“Common” is measured: kept by at least 6 in 10 of the 11 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Google Cloud CLI

Use gcloud for bounded account, project, IAM, API, billing, logging, and audit operations.

Defaults

Google Cloud is the default cloud for Fmind projects. Resolve the project and region from explicit project configuration, or ask when unset. Prefer Cloud Run for services; use GKE only when the project explicitly adopts Kubernetes. Existing project choices take precedence.

Workflow

  1. Resolve identity and scope: inspect the named configuration, account, project, and billing project; never activate another configuration just to make a command work.

    gcloud config configurations list --format='table(name,is_active)'
    gcloud config configurations describe <configuration> --format=json
    gcloud config get auth/impersonate_service_account --configuration <configuration>
    gcloud config get auth/access_token_file --configuration <configuration>
    gcloud projects describe <project-id> --format='json(projectId,projectNumber,lifecycleState)'
  2. Pin every consequential call: pass --configuration, --account, --project, and --billing-project (plus --impersonate-service-account for an approved chain) so terminal defaults cannot redirect the operation.
  3. Start read-only and bounded: use supported --filter and --limit on listings and selected fields in --format; for logs, constrain time and resource before projecting message fields. Keep full IAM policies or configuration when required for the review; a limited listing is not an exhaustive audit. --quiet controls prompts, not output volume.
  4. Authorize mutations: API enablement, IAM, billing, deletion, and production changes need explicit authority.

Gotchas

  • --quiet is not authority: it suppresses prompts; it neither makes an operation safe nor approves it.
  • Overrides replace the account: CLOUDSDK_AUTH_IMPERSONATE_SERVICE_ACCOUNT, CLOUDSDK_AUTH_ACCESS_TOKEN_FILE, and --impersonate-service-account change the effective principal; prove which is active first.
  • Firebase reuses ADC: the global firebase CLI (firebase-tools) authenticates with gcloud Application Default Credentials, so skip firebase login; pass --project <id> and --json, and treat deploy like any mutation.
  • Failures are findings: on an auth, permission, or quota error report the missing principal, denied permission, or exhausted quota; do not switch configurations, broaden IAM, create keys, or move spend to another project.

Official Skills

Upstream: google/skills (skills/cloud), listed and installed through Google catalog; its CLI guardrail skill applies to every gcloud call. That skill is also named gcloud: preview it and apply its guardrails, but never install it under that name (vendor-skill policy).

Documentation

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.