antigravity CLAUDE.md

antigravity CLAUDE.md is an instructions file for Claude Code from JamesZor/antigravity. It costs 1,768 tokens per session, scanned A, original, MIT.

Instructions for using an MCP server that lets one AI agent delegate bulky work to Google's Gemini agent.

In plain words
What is it for?
It supports workflows for web research, codebase indexing, commit-message generation, cross-model review, architecture work, and other delegated coding tasks.
Why use it?
Delegating tasks such as research or codebase indexing can keep the main agent's working context smaller and leave it focused on coordinating the work.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md; mentions Claude Code.

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 instructions/jameszor/antigravity/claude-md
Clone the repo
git clone --depth 1 https://github.com/JamesZor/antigravity

Made for: Claude Code.

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.

agentmods badge for antigravity CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/jameszor/antigravity/claude-md.svg)](https://agentmods.dev/instructions/jameszor/antigravity/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/jameszor/antigravity/claude-md"><img src="https://agentmods.dev/badge/instructions/jameszor/antigravity/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,768 This file is loaded in full into every session.
When invoked 1,768 The same file — it is already loaded in full.
Security scan A 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.01768 $0.01768
Opus 5 $0.00884 $0.00884
Sonnet 5 $0.00354 $0.00354
Haiku 4.5 $0.00177 $0.00177

Measured 5d ago against content hash 26976149e8bf, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

antigravity CLAUDE.md scanned grade A 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 5d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **SKILL.md files use relative script paths** (`uv run scripts/x.py`), but a worker's cwd is its own job dir, not the skill dir. `skill_preamble()` therefore injects each skill's **absolute** path. Without it, every scr
CLAUDE.md · 58 lines

How it starts

The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

What this is

An MCP (Model Context Protocol) server that exposes the Antigravity CLI (agy, Google's Gemini agent) as a set of tools usable from Claude Code, Claude Desktop, Cursor, and Windsurf. The animating idea: Claude is the Conductor/frontier model, and agy/Gemini is a cheaper executing agent to which Claude offloads bulky, token-heavy work (web research, codebase indexing, commit-message generation, cross-model review) to keep Claude's own context lean. See the antigravity_workflow and antigravity_research_recipe MCP prompts (in antigravity_mcp/prompts.py) for the operating philosophy.

The implementation lives in the antigravity_mcp/ package; main.py is a thin entrypoint. Layout: app.py (the shared FastMCP instance), agy.py (subprocess/model-tier helpers + the pure skill layer, no MCP — unit-testable), jobs.py (filesystem-backed background jobs + _start_batch fan-out scaffolding), delegation.py, research.py, architect.py, skills.py, codegit.py, prompts.py, and server.py (side-effect imports that register everything + main()).

Three multi-step pipelines build on this: deep-research (research_fanoutresearch_statuscollect_digests, plus the /deep-research skill and antigravity_research_recipe prompt); architect/build (propose_design_questionsreview_fanout/research_fanoutdraft_design_doc, plus the /architect skill and antigravity_build_recipe prompt) — a Spec-Driven loop (Requirements → Design → Tasks → Implement) for creating/improving codebases; and science (propose_science_planscience_fanoutresearch_statuscollect_digests, plus the /science-research skill and antigravity_science_recipe prompt). All three share the same fan-out scaffolding (_start_batch) and batch collectors (research_status/collect_digests are batch-generic — they read any sub_NN/subreport.md + manifest.json, so review and science batches reuse them unchanged).

Read the full file on GitHub · 58 lines

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.

  1. 5d ago First seen · 58 lines · 1,768 tokens per session scan A 26976149e8bf

Subscribe to this mod's changes

antigravity CLAUDE.md is an instructions file published in the GitHub repository JamesZor/antigravity (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,768 tokens to every session, about $0.0088 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other instructions, from other repositories