- Home
- Skills
- Agents & Orchestration
- claude2antigravity
More screenshots
Works with the AI Agents you already use
claude2antigravity
Keep Google Antigravity in sync with Claude Code 24/7 - skills, plugins, agents, MCP servers and conversations - and continue your work the moment Claude hits its usage limit.
$3.99
claude2antigravity
Example session with this skill installed
Claude just told me I hit my usage limit while I was fixing the login bug in my web app. Get Antigravity ready so I can continue right now: make sure it has my Claude skills and plugins, and give me a handoff brief of where I stopped.
- Read your context and instructions
- Compiled the claude2antigravity
Done - Antigravity is ready, you can continue right now.
The live sync was already running, so everything from Claude is in Antigravity. I wrote a handoff brief of where you stopped.
What is in Antigravity
| Item | Status |
|---|---|
| Skills | 140 synced (2 new today: api-tester, release-notes) |
| Plugins | 31 synced, converted and enabled |
| Agents | 1 synced (code-reviewer) |
| MCP servers | 2 added; your own servers were kept |
| Conversations | 24 Claude sessions, readable in Antigravity |
Handoff brief
Saved to ~/.gemini/config/memory/claude2antigravity/HANDOFF.md
- Project:
C:/work/shop-app - Branch:
fix/login-redirect - Claude stopped: 14:32 UTC (usage limit)
- Changed files:
src/auth/session.ts,tests/auth.test.ts - Your last request: "Fix the redirect loop after login and add a test for it."
Todo list
- Find why
/loginredirects to itself - Fix the session check in
session.ts - Add a test for the redirect
- Run the test suite
Next steps
- Open Antigravity and start a new conversation.
- Say:
continue from the Claude handoff. Antigravity reads the brief and starts with the first open todo (the redirect test).
3. When Claude is back, keep working there. New skills, plugins and agents keep syncing automatically.
The old result table had 4 columns, and they were cut off in the preview ("Ready in memory/claude…"). The 2-column version fits.
Connects securely to your tools. The creator never sees your data.
About this skill
Hit your Claude usage limit? Open Antigravity and keep working - with everything you had in Claude.
You build up a lot in Claude Code: skills, plugins, agents, MCP servers and hours of conversation about your project. Google Antigravity can keep going when Claude has to wait, but it knows none of it. Claude2Antigravity Pro fixes that, automatically and in the background.
What it does
- Live sync, about 1-2 seconds. Install or edit a skill, plugin or agent in Claude and it is in Antigravity moments later, converted to Antigravity's own formats (plugin manifests, MCP config, agent files, hooks). A small notification tells you what arrived.
- Usage-limit handoff. When Claude says "usage limit reached", Pro writes a "continue here" brief - project folder, git status, latest plan, todo list and last messages - and Antigravity picks it up in your next request. Just say: continue from the Claude handoff.
- Your Claude work in every Antigravity request. A light hook adds your latest Claude session to each request, so Antigravity always knows what you were doing.
- Claude Desktop chats too, read offline from the Desktop app, kept even after the app cleans up.
- Search everything:
c2a.py search <words>finds any message in all your past Claude conversations. - Privacy filter: e-mail addresses, API keys and tokens are replaced by
[hidden]in synced text.
Safe by design
- Only items it installed are ever updated or removed; your own Antigravity items are never touched. An edited copy is kept.
- Your
CLAUDE.mdis never modified. sync --dry-runpreviews every change;uninstallremoves everything it added, including the auto-start task.- No network access at all, no environment variables, no secrets - everything stays on your computer.
Light and fast (measured, numbers in the GitHub README)
- Background check: under 1 ms per second, about 20-35 MB of memory, no disk writes while nothing changes.
- New skill in Claude -> in Antigravity: about 1.2-1.4 s median.
Setup: one command
python <skill folder>/scripts/c2a.py setup
It installs the program, adds the Antigravity hook, runs the first sync and starts the live sync at login (Windows Task Scheduler, macOS LaunchAgent or Linux systemd). Prefer manual control? Use setup --no-autostart.
Free vs Pro The free Community edition on GitHub copies everything when you run it by hand. Pro adds the 24/7 live sync, notifications, auto-start, the usage-limit handoff, live context in every request, Claude Desktop chats, search and the privacy filter.
Use cases (if the form asks)
- Claude hit its usage limit in the middle of a task - continue in Antigravity from the handoff brief.
- Use the same skills, plugins and agents in Claude Code and Antigravity without copying anything by hand.
- Find something you discussed with Claude weeks ago (search across every conversation).
- Work in both tools on one project and keep them in step automatically.
Permissions (what to state if asked)
- Reads: ~/.claude, ~/.claude.json, the Claude Desktop app data folder.
- Writes: ~/.gemini/config (skills, plugins, agents, mcp_config.json, plugin list, one marked block in GEMINI.md) and ~/.claude2antigravity.
- Runs: Task Scheduler / launchctl / systemctl (auto-start, at setup, with the user's consent),
wsl.exe -l(find Claude inside WSL),git status(handoff brief), its notification window. Never through a shell, never with admin rights. - Network: none.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
- Needs Python 3.10+ on the computer.
- Antigravity shows new skills in a new conversation; new plugins and MCP servers may need an Antigravity restart.
- Claude Desktop chats are read from the app's local storage; if that format changes, Desktop import may pause until an update (Claude Code sync is not affected).
- Items already in Antigravity with the same name are not overwritten unless you choose --adopt.
- Sync goes from Claude to Antigravity; the reverse direction is on the roadmap.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 4 days ago
Windows only
- Software
Python 3.10+
Windows, macOSorLinuxClaude CodeGoogle Antigravity- Optional
Git
Permissions
File Scopes
Reads Claude's own folders (skills, plugins, agents, conversations, user MCP servers) and writes them into Antigravity's config folder in Antigravity's format. Its own state and logs are kept in ~/.claude2antigravity. The terminal is used only to register auto-start (Task Scheduler / launchctl / systemd, at setup, with the user's consent), to find Claude inside running WSL distros (wsl -l) and to add git status to the handoff brief. It never runs through a shell, never needs admin rights and makes no network calls.