A command for saving the default Gemini model and Google Cloud region used by the worker. The settings are stored for future sessions and can use either Vertex AI credentials or a Gemini API key.

Use it to select a supported Gemini model, choose a Vertex AI location, and set the authentication mode for later sessions.

It avoids choosing the same model, region, and connection method every time a session starts, while rejecting unsupported model names before work begins.

setup — Command for coding agents by jinseo-jang: A command for saving the default Gemini model and…

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add commands/jinseo-jang/antigravity-plugin-cc/setup

Clone the repo

git clone --depth 1 https://github.com/jinseo-jang/antigravity-plugin-cc

Or install agy, the plugin that ships this one along with the rest of its 11 commands, 2 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/commands/jinseo-jang/antigravity-plugin-cc/setup.svg)](https://agentmods.dev/commands/jinseo-jang/antigravity-plugin-cc/setup)

Your own site

<a href="https://agentmods.dev/commands/jinseo-jang/antigravity-plugin-cc/setup"><img src="https://agentmods.dev/badge/commands/jinseo-jang/antigravity-plugin-cc/setup.svg" alt="Measured on agentmods" height="20"></a>

Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 968 The whole file, excluding the scripts and references it only reads on demand.

Security scan B 1 finding. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00019	$0.00968
Opus 5	$0.00010	$0.00484
Sonnet 5	$0.00004	$0.00194
Haiku 4.5	$0.00002	$0.00097

Measured 4d ago against content hash 01d14bc07e35, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

setup scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Asks for rootmediumPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

- `--api-key` is never accepted or stored. Provide the key one of three ways, checked in this order: (1) **OS keychain** — `python -m keyring set cao gemini_api_key` (encrypted at rest, recommended); (2) **`GEMINI_API_KE

plugin/commands/setup.md · 42 lines

What it actually says

Persist model/region defaults to $CAO_PLUGIN_DATA/defaults.json (or ~/.config/cao/defaults.json). These are read fresh on every session start — no daemon restart needed.

Supported models:

Model	Status	Notes
`gemini-3.7-flash`	GA	Default.
`gemini-3.6-flash`	GA
`gemini-3.5-flash`	GA
`gemini-3.5-flash-lite`	GA
`gemini-3.1-pro-preview`	Public Preview	Narrower regional availability.

Any other model string is rejected with -32602 before the worker starts. Region is your choice — global (the default) works for all models; pick any Vertex location where your model is available (Gemini's regional coverage keeps expanding).

Workflow:

Use AskUserQuestion to ask the user which mode they want: vertex (Vertex AI via ADC) or gemini_api_key (Gemini API key).
For vertex mode, ask for --model (gemini-3.7-flash (default), gemini-3.6-flash, gemini-3.5-flash, gemini-3.5-flash-lite, or gemini-3.1-pro-preview) AND --location — recommend global (today the only region these Gemini-3 models are reliably served on); let the user type another Vertex region only if they know their model is available there (an unavailable region hangs until the worker-turn timeout, it does not fail fast). If they choose nothing, global is used. Also ask for the GCP project id — tell the user they can leave it blank to let agy auto-detect it from ADC (quota_project_id) or the active gcloud config. For gemini_api_key mode, ask for --model only (key stays in GEMINI_API_KEY; no location or project).
Run the companion, including --project <id> only if the user gave one (omit it entirely to let auto-detect handle it):

!python "${CLAUDE_PLUGIN_ROOT}/scripts/cao-companion.py" --plugin-data "${CLAUDE_PLUGIN_DATA}" setup --mode [--model ] [--location ] [--project ]

If the companion prints a rejection (unsupported model), relay the Options from the message via AskUserQuestion and retry with a corrected model.
On success, confirm to the user that the defaults are saved and will apply to the next /agy:implement invocation.

Notes:

--api-key is never accepted or stored. Provide the key one of three ways, checked in this order: (1) OS keychain — python -m keyring set cao gemini_api_key (encrypted at rest, recommended); (2) GEMINI_API_KEY env var — export it in the shell that launches Claude Code, then restart it (Google's recommended location); (3) plaintext file ~/.config/cao/gemini_api_key (chmod 600) — last resort, not encrypted.
If the companion warns that no key is found (gemini_api_key mode), point the user at the keychain command first, then the env-var or file fallbacks.
Region is user-selectable. global is the default (works for all supported models) and is used when you omit --location; pick another Vertex region where your model is served. An unavailable model×region is not blocked at setup — it surfaces as a worker timeout, so choose a region where the model actually runs.
gemini-3.1-pro-preview is the correct model code for Gemini 3.1 Pro (bare gemini-3.1-pro is a 404).
Defaults take effect on the next session (fresh-read, no daemon restart required).
To clear defaults, delete defaults.json directly.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

4d ago First seen · 42 lines · 19 tokens per session scan B 01d14bc07e35

Subscribe to this mod's changes

setup is a command published in the GitHub repository jinseo-jang/antigravity-plugin-cc (6 stars, last pushed 21d ago), licensed Apache-2.0. It adds 19 tokens to every session and 968 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

setup

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade B, and why

Asks for rootmediumPrivilege escalation

What it actually says

What this file has done since we first saw it

Other commands, from other repositories

git

checklist

clarify

specify

converge

implement