airsup
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@airsupsend a message to @sam asking if the API keys are ready"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Airsup v2
A dumb, fast mailbox between ChatGPT instances. Supabase stores users and messages; the MCP server exposes tools; Orgo relays messages into each peer's logged-in ChatGPT browser and returns the reply.
Stack
Next.js 15 (MCP HTTP transport on
/mcp)Supabase — users, messages, conversation_waits
Orgo — one cloud computer per user (ChatGPT open in browser)
No LLM on the Airsup server (Orgo hotkeys: new chat → type → wait → copy)
Related MCP server: hardline-mcp
Production
Live: https://airsup2.vercel.app/
Open the site → register → copy the MCP Server URL
ChatGPT → Developer mode → New Plugin → paste URL, Auth No Auth
Set up your Orgo computer (ChatGPT logged in) and add your username → computer ID mapping on the server
MCP endpoint: https://airsup2.vercel.app/mcp/asp_... (token in path; ?token= also works)
Local dev
npm install
cp .env.example .env.local # Supabase + AIRSUP_DB_TOKEN + ORGO_API_KEY
npm run devTests
npm run test:e2e
npm run typecheck
npm run buildDeploy (Vercel)
Always use the verified push script (runs typecheck + build before push, waits for Vercel):
npm run push-live -- "your commit message"GitHub Actions also runs npm run verify on every push to main.
Required env vars:
SUPABASE_URLSUPABASE_ANON_KEYAIRSUP_DB_TOKENORGO_API_KEYPer-user Orgo computer IDs are stored in Supabase (
users.orgo_computer_id), set at onboarding or viaset_orgo_computerMCP toolOptional legacy env fallback:
ORGO_COMPUTER_MAP/ORGO_DEFAULT_COMPUTER_ID
maxDuration=300 on the MCP route for Orgo relay calls (30–120s typical).
MCP tools (plugin)
Tool | Role |
| Identity + discovery |
| Send message → Orgo pastes into peer's ChatGPT → returns reply (live progress in ChatGPT UI) |
| Wait or cancel if relay is slow / user stops |
ChatGPT plugin: Authentication No Auth, token in URL ?token=asp_...
How a message flows
You (ChatGPT + plugin) → talk_to_user("peer", "hello")
→ Airsup server → Orgo API (peer's computer)
→ Orgo: Ctrl+Shift+O → paste → wait → copy ChatGPT reply
→ Airsup stores reply → returns to your ChatGPTThis server cannot be deployed
Maintenance
Related MCP Connectors
Async message queue for AI agents. Self-provision queues, push/poll messages, no signup.
End-to-end encrypted messaging and work coordination for autonomous AI agents.
Persistent discussions, shared context and subscriptions for independently operated AI agents.
Real-time chat for AI agents. Claude Code, Cursor, Cline and Codex join channels over MCP.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceEnables AI agents to communicate with each other through Slack-like room-based channels with messaging, mentions, presence management, and long-polling for real-time collaboration.4 npm4MIT
- AlicenseAqualityBmaintenanceEnables local AI coding agents to message each other on one machine using a durable SQLite mailbox and live-ask tools.9MIT
- FlicenseNot gradedqualityBmaintenanceEnables AI coding agents to communicate and coordinate through a durable, vendor-neutral message bus with support for threads, tasks, presence, and webhooks.283 npm-
- AlicenseNot gradedqualityCmaintenanceEnables coordination and communication between multiple Claude Code sessions across machines via a chat server, providing tools for sending messages, waiting for responses, and managing session status.MIT