Antigravity CLI in NeuroSquadAntigravity CLI,with a canvas around it
Google’s terminal agent — the successor to Gemini CLI — in as many cards as you need, each with a real terminal. NeuroSquad knows when each one is working, waiting on you or finished, hands it terminals, browsers and notes by arrow — and the new tools arrive while it runs.
Official siteantigravity.google/product/antigravity-cli- The real
agyCLI, your own Google sign-in - Nothing written to your
~/.gemini - Free, for Windows and macOS
It knows when agy is done — and when it’s asking
A quiet terminal can mean “finished” or “waiting for your confirmation”. NeuroSquad doesn’t guess: every Antigravity card reports its turns through agy’s own lifecycle hooks and the state agy hands its window-title command.
PreInvocationwhat agy sentYou press Enter and agy fires PreInvocation before it calls the model. The card’s icon spins, the Squad status card counts it as working, and the toast from its last question is taken down.
tool_confirmation_pendingwhat agy sentagy opens a confirmation and its state says tool_confirmation_pending. The card pulses, and the toast names what it wants to run — the command or tool its PreToolUse hook reported just before.
Stopwhat agy sentStop closes the turn. The next queued prompt goes out, the journal gets the answer from agy’s transcript, and a toast says it’s done. A turn you interrupt with Esc ends the same way.
Its questions too
When agy asks you something with ask_question, the question itself comes to the card and the toast.
One status everywhere
The card, the sidebar, the Squad status card, custom cards and the phone all read the same status.
Your hooks stay yours
NeuroSquad’s hooks are added to the card’s own hooks.json, next to yours; the hooks in your ~/.gemini/config keep running.
Three things you’ll do every day
Replicas of the app, played step by step. When agy asks for confirmation, the answer is yours.
Antigravity CLI fixes a failing test through a connected terminal, stops to ask before deleting a folder, and finishes — the journal writes itself.
Everything Antigravity CLI can do in NeuroSquad
Antigravity CLI is integrated as fully as Claude Code. Here is all of it, with a docs page for each — and, at the end, what it honestly doesn’t get.
- 01
Knows its state
agy’s own hooks and its state report every turn, so nothing is guessed.
Finished and waiting are different
Each card gets agy’s lifecycle hooks:
Docs: Finished and waiting are differentPreInvocationmarks it working,Stop— finished,PreToolUseremembers the tool it is about to use. Its title command’s state adds the confirmation on screen, the context and the quota.A notification that says what it needs
The toast leads with what agy wants to do —
Docs: A notification that says what it needsrun_command — rm -rf dist, or the tool it calls. One per agent; a waiting one stays until you answer.Every card is its own conversation
agy picks the conversation id; the card learns it from the hooks and resumes it with
Docs: Every card is its own conversation--conversation <id>after a restart./clearmoves the card to the new conversation.The Squad status card
Every AI agent of the workspace on one card, with its status, the live time of its turn and its model — the one agy reports to its hooks.
Docs: The Squad status cardCustom cards can follow it
Cards built with the Card SDK see the same statuses, turn events and last replies, and
Docs: Custom cards can follow itagent_waitgets an Antigravity helper’s answer.
- 02
Works through the canvas
An arrow from the card is a set of tools. Draw it, and agy can use what is at the other end — right away.
Arrows are tools
An arrow to a browser, terminal, note or another agent gives agy
Docs: Arrows are toolsbrowser_*,terminal_*,note_*oragent_*from NeuroSquad’s MCP server — allowed by amcp(neurosquad/*)rule, so it never asks. Each call lights up its arrow.New arrows work mid-session
agy re-lists a server’s tools when they change, so an arrow drawn while it works adds the tools to its very next request. No restart, no new conversation.
Docs: New arrows work mid-sessionIt can add cards itself
Docs: It can add cards itselfcanvas_spawn_cardlets it open a terminal, a browser, a note or a helper agent — Antigravity or any other CLI — next to itself, already joined by an arrow.Canvas mode
A
Docs: Canvas modePreToolUsehook denies agy’s own shell and web tools (run_command,manage_task,search_web,read_url_content, the browser), so commands run in terminal cards you can watch. The denial holds in dangerous mode too.Skills from skills.sh
An arrow to a skill card gives agy
Docs: Skills from skills.shskill_loadwith each skill and when to use it — and like every arrow, it works without a restart.Plugin cards on its turns
Memory, Context7, Caveman and Graphify reach agy’s turns as a short note it reads before the model call — through
Docs: Plugin cards on its turnsinjectStepsof itsPreInvocationhook.
- 03
Built for long runs
Hours-long sessions run into full contexts and quotas. Both are handled on the card.
Context meter
A small ring in the header shows how full the context is, against the window agy reports for its model. agy compacts the context by itself when it fills up.
Docs: Context meterContinues after a quota
agy retries short rate limits itself. When a turn ends on a daily quota, the card shows the limit with its reset time and sends “continue” then.
Docs: Continues after a quotaPrompt queue
Line up the next messages: each goes out on
Docs: Prompt queueStop, when the turn is really over — never in a pause mid-thought. Daily prompts too.Hand off to a fresh agent
When a session is full, start a new agent beside it — Antigravity or any other CLI — with a summary of the work typed in, ready for you to send.
Docs: Hand off to a fresh agentJournal
Each finished answer is added to a connected note, read from agy’s own transcript: a running record of what the agent did.
Docs: Journal
- 04
Works in a team
Several Antigravity agents — or a mix of CLIs — on one project, without touching your own agy setup.
Your
~/.geministays untouchedEach card runs with a Gemini dir of its own (
Docs: Your ~/.gemini stays untouched--gemini_dir): its settings and config layered over yours, your conversations, onboarding, plugins, skills and rules linked. Nothing of yours is written.Isolated worktrees
Give an agent its own git worktree, so parallel agents never edit the same files. A setup command and files to copy, like
Docs: Isolated worktrees.env, prepare it before agy starts.Add a reviewer
One click starts an agent on a different CLI in the same worktree, with an arrow from the author and a prompt to review without editing anything.
Docs: Add a reviewerDangerous mode, per card
For a run you trust, a card restarts agy with
Docs: Dangerous mode, per card--dangerously-skip-permissionsand resumes the same conversation — agy has no way to switch it live. A per-card switch, off by default.One instructions file
agy reads AGENTS.md itself (and GEMINI.md), so the AGENTS.md you edit in NeuroSquad reaches it as is — one file for every CLI on the canvas.
Docs: One instructions file
- 05
Cost, control and access
What it used, what it may spend, which model it runs on, and how you reach it.
Tokens per card
Every request’s tokens — input, cached, output — read from agy’s own transcript, subagents (their own conversations) folded into their parent, split exactly by card.
Docs: Tokens per cardBudget brake
A limit per workspace. Past it, each agent’s turn is interrupted with Esc — the conversation is kept — and automatic prompts stop. Raise the limit and it’s lifted.
Docs: Budget brakeOpenRouter models
Run a card on any OpenRouter model, or on your own OpenAI-compatible provider. agy speaks the Gemini API, so the card talks to a small gateway inside NeuroSquad through agy’s API-key mode; your key never reaches agy.
Docs: OpenRouter modelsMCP servers from the registry
Install a server from the official MCP Registry once, then give it to an agent with an arrow. It isn’t pre-allowed: agy asks before every call to it.
Docs: MCP servers from the registryFrom your phone
Scan a QR code and the whole canvas opens in the phone’s browser: read agy’s terminal, answer its confirmation, send the next prompt.
Docs: From your phone
- 06
Not available for this CLI
What Antigravity CLI does not expose, or the app does differently for it. Said plainly.
Closed source
agy is proprietary: its repository holds the README and the changelog, not the code. Everything here was measured against its behaviour (version 1.3.1), not read from its source.
Docs: Closed sourceSide calls aren’t counted
agy titles each conversation with a small extra model call that its transcript doesn’t record. Usage shows the conversation’s requests, exactly.
Docs: Side calls aren’t countedNo price for Google’s own models
There are no Gemini list prices in NeuroSquad yet, so cards on your Google sign-in show tokens and “no price” — never a made-up $0. OpenRouter cards are priced from its catalogue.
Docs: No price for Google’s own modelsHooks cost a moment
agy’s hooks are processes it waits for, so only the events a status needs are used: the model calls and the end of a turn, plus the one tool event.
Docs: Hooks cost a momentGemini-only features stay on Google
Through the OpenRouter gateway, what OpenAI-style chat has no equivalent for — Google Search grounding, thought signatures — does not survive.
Docs: Gemini-only features stay on GoogleNo RTK
RTK rewrites shell commands before they run. agy’s
Docs: No RTKPreToolUsehook can allow, ask or deny a command, but not change it — so RTK can’t work with it.
Who’s waiting, and what it used
Two cards that answer the questions you ask most when several Antigravity agents run at once.
The Squad status card
Every agent of the workspace with its status, model and turn time. The one waiting on you comes first, with what it asks. Tap a status to filter.
- coupon-fixAntigravity CLIgemini-3.1-pro-lowAntigravity CLI needs your permission: run_command — rm -rf distNeeds you2:34
- api-migrationAntigravity CLIgemini-3.1-pro-highMove routes to the new clientWorking11:07
- lib-researchAntigravity CLIgemini-3.8-flash-mediumCompare two date librariesWorking5:25
- docs-passAntigravity CLIgemini-3.8-flash-lowUpdate the READMEFinished3 min. ago
Usage & cost
Read from agy’s own transcripts on your disk and split by card. The rows always add up to the total.
| Agent | Tokens | Cost | ||
|---|---|---|---|---|
| coupon-fix | 912,406 | $2.14 | ||
| api-migration | 604,118 | $1.39 | ||
| docs-pass | 188,530 | $0.09 | ||
| Total | 79 | 1,705,054 | $3.62 |
The real Antigravity CLI, plus a Gemini dir of its own
NeuroSquad starts the same `agy` you run yourself, in a real terminal in the card. This is everything it adds:
Nothing written to your ~/.gemini
The card’s settings, hooks.json and mcp_config.json are its own; your conversations, plugins, skills and rules are linked. Deleting a card unlinks before it removes.
Your sign-in, your plan
agy runs with the Google sign-in or the Gemini API key it already uses. Its first-run screens — the colour scheme and Google’s terms and data-use consent — are yours to answer; NeuroSquad never answers them for you.
A conversation is your conversation
Because the card’s Gemini dir links your real conversations, agy --conversation <id> works from a plain terminal too — and the other way round. agy’s self-update is off inside cards.
Antigravity CLI in NeuroSquad, answered
Give Antigravity CLI a canvas
Free for Windows and macOS. Your code, your sign-in and your usage stay on your machine.