ariekogan/ateam-mcp

**Give any AI the ability to build, validate, and deploy production multi-agent systems.**
This is an MCP server that connects AI assistants — ChatGPT, Claude, Gemini, Copilot, Cursor, Windsurf, and any MCP-compatible environment — directly to the ADAS platform.
An AI developer says *"Build me a customer support system with order tracking and escalation"* — and their AI assistant handles the entire lifecycle: reads the spec, builds skill definitions, validates them, deploys to production, and verifies health. No manual JSON authoring, no docs reading, no copy-paste workflows.
Why this matters
Today, building multi-agent systems requires deep platform knowledge, manual configuration, and switching between docs, editors, and dashboards. **ateam-mcp eliminates all of that** by making the ADAS platform a native capability of the AI tools developers already use.
The AI assistant becomes the developer interface:
No context switching. No manual steps. The full ADAS platform — specs, validation, deployment, monitoring — is available as natural language.
How it reaches the AI community
ChatGPT users
ChatGPT supports MCP connectors in Developer Mode. Users connect by pasting a single URL:
**Settings → Connectors → Developer Mode → paste https://mcp.ateam-ai.com**
That's it. All 12 ADAS tools appear in ChatGPT. Any ChatGPT Pro, Plus, Business, or Enterprise user can build and deploy multi-agent solutions through conversation.
Claude users
**claude.ai and Claude Desktop** — add the hosted connector: open Customize > Connectors (Claude Desktop: Customize in the sidebar, then Connectors), click **Add custom connector**, enter https://mcp.ateam-ai.com, and click **Connect**. The A-Team sign-in page opens in the browser and asks for your workspace's key: copy it from app.ateam-ai.com/connect (a workspace owner or admin) and paste it there. The key decides the workspace; to switch, Disconnect and Connect again with the other workspace's key.
**Claude Code** — the hosted connector, signed in the same way:
then /mcp and authenticate in the browser.
Never paste a key into a chat. ateam_bootstrap's session field says which workspace a session is on and how to switch.
Cursor / Windsurf / VS Code (Copilot)
Add the hosted server to .cursor/mcp.json, mcp_config.json, or .vscode/mcp.json:
A client with OAuth opens the A-Team sign-in page; one without sends the key as Authorization: Bearer from its own config (see below). A local process (npx -y @ateam-ai/mcp) signs in only through ateam_auth: an ADAS_API_KEY in its environment does not sign it in.
Gemini and other platforms
As MCP adoption grows (it's now governed by the Agentic AI Foundation under the Linux Foundation, co-founded by Anthropic, OpenAI, and Block), every AI platform that implements MCP gets access to ateam-mcp automatically. The remote HTTP endpoint (https://mcp.ateam-ai.com, also at /mcp) works with any client that supports Streamable HTTP transport.
The HTTP endpoint needs a credential on every request: OAuth (clients that support it follow the 401 challenge on their own), or your A-Team API key sent as Authorization: Bearer . An anonymous request is refused with 401; since 0.4.93 there is no anonymous session to call ateam_auth from. See CHANGELOG.md.