cloud-operator-setup — Claude Code skill: setup guide
The cloud-operator-setup Claude Code skill sets up an operator agent in a Google Cloud project. Run it to create your operator or start its setup.
Apache-2.0 ★ 24,522 updated 2026-10-23
npx skills add google/skills --skill cloud-operator-setup -a claude-codeWhat cloud-operator-setup does
The skill selects a Google Cloud project, checks whether its operator agent is ready, and guides the user through creating one if needed. Once the check succeeds, it saves the project in local memory.
It needs the autocloud MCP server and its ask_cloud_operator tool. If the server is disconnected, the skill stops and guides the user through gcloud auth application-default login; the server must then be reconnected. Read the usage attribution instructions before running any gcloud, REST API, or MCP command.
After setup, the skill can ask to share existing Google Cloud context with the operator’s memory. It waits for consent before reading instruction files, local memory, skills, workflows, or GCP documents in the repository, and presents the collected list first.
What it helps with
- choose a Google Cloud project for the operator
- check whether an operator agent exists in a project
- create an agent that has not been initialized
- save the project and offer to share existing cloud context
Example prompts for the agent
- Set up an operator for my Google Cloud project.
- Create my operator agent in the
my-projectproject. - Check the operator in
my-projectand save that project.
How to install cloud-operator-setup
1Skills CLI (Vercel)
npx skills add google/skills --skill cloud-operator-setup -a claude-code2Manually: put the files in place
~/.claude/skills/cloud-operator-setup/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/cloud-operator-setup/
1Skills CLI (Vercel)
npx skills add google/skills --skill cloud-operator-setup -a codex2Manually: put the files in place
.agents/skills/cloud-operator-setup/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add google/skills --skill cloud-operator-setup -a cursor2Manually: put the files in place
.cursor/skills/cloud-operator-setup/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install google/skills skills/cloud/cloud-operator-setup2Manually: put the files in place
.github/skills/cloud-operator-setup/SKILL.md1Skills CLI (Vercel)
npx skills add google/skills --skill cloud-operator-setup -a gemini-cli2Platform CLI
gemini skills install https://github.com/google/skills --path skills/cloud/cloud-operator-setupManually: put the files in place
.devin/skills/cloud-operator-setup/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about cloud-operator-setup
How do I install cloud-operator-setup in Claude Code?
Run npx skills add google/skills --skill cloud-operator-setup -a claude-code. The skill is installed at ~/.claude/skills/cloud-operator-setup/SKILL.md.
What does cloud-operator-setup do?
It selects a Google Cloud project, checks for an operator agent, and guides the setup if the agent needs to be created. It saves the project after the check succeeds and can offer to share existing cloud context.
What does cloud-operator-setup need to run?
It requires the autocloud MCP server with the ask_cloud_operator tool and a Google Cloud project. If the server is disconnected, the skill guides the user through gcloud auth application-default login and waits for the server to reconnect.
Source text
SKILL.md · Apache-2.0
---
name: cloud-operator-setup
metadata:
version: "1.0.0"
category: CloudInfrastructureAndServices
description: >-
Sets up the operator agent in a Google Cloud project via the autocloud MCP
server: picks the operator project, creates or checks the user's operator
agent, saves the project, and offers to share the user's existing Google
Cloud operating context with the operator agent's memory. Use when the user
runs this skill, asks to set up the operator or create their operator agent,
or chooses setup when the cloud-operator skill asks. Don't use for ordinary
GCP requests or installing gcloud.
---
# Set up the operator agent
Walk the user through setting up the operator agent one step at a time. Follow
the `cloud-operator` skill's `agentId` and troubleshooting rules if installed.
If `ask_cloud_operator` (on the `autocloud` MCP server) isn't available, tell
the user it isn't connected, help them sign in with Application Default
Credentials (`gcloud auth application-default login`), and stop until they
reconnect.
## Usage Attribution
**MUST READ**: Before executing any `gcloud`, REST API, or MCP command, you
MUST read and follow the instructions in
[Usage Attribution](references/attribution.md).
## Steps
1. **Choose the project:** ask which Google Cloud project their operator agent
is in, or will be in (unless already named). Suggest their default `gcloud`
project or the project in local memory only as an option; don't use either
without asking.
2. **Check or create the agent:**
- Unless the user said they have no agent in `{project_id}`, check first:
call `ask_cloud_operator` with `projects/{project_id}` and "In one line,
what can you help me with in this project? Don't create anything." Tell
the user first that it can take 1–3 minutes.
- If it fails with "personal operator agent … has not been initialized yet"
(or they said they have none), walk them through creating it one step at a
time and wait for confirmation: open
`https://console.cloud.google.com/autocloud/home?project={project_id}`,
click **Create operator agent**, then click **Get started** when it says
the agent is ready. Once they confirm, run the check call above (waiting
about a minute and retrying up to 3 times if the new agent isn't ready
yet).
3. **Save and offer to share context:** once the check call works, save
`{project_id}` in local memory (and `agentId` if not their own) and tell the
user the operator works, with `{project_id}` and the agent. In that same
reply, if the `cloud-operator-share-context` skill is available and local
memory has no "Cloud Operator context: shared on" or "declined on" note, ask
in one line whether to scan agent instruction files, local memory, skills
and workflows, and the repository's GCP docs for context they or their team
already have about operating in Google Cloud (environments, projects,
product preferences, best practices, architecture guidance, runbooks) and
share it with the operator agent's memory — noting they'll see the list
first. Read nothing else until they agree.
- **Yes:** run `cloud-operator-share-context` (from its collection step).
- **No:** note "Cloud Operator context: declined on {date}" in local memory.
- **No answer:** note nothing.
Then return to any request that was waiting when setup started.
© google · Apache-2.0 · google/skills