Skip to content
BuildBase
Agent-ready apps

Let your users connect Gemini CLI to your app

Agent-ready apps: make the app you build with the SDK ready for Gemini CLI. Your users add your app's MCP server to Gemini CLI, sign in on your hosted page and approve it, and Gemini CLI acts as them, never as more.

How your users add your app to Gemini CLI

Give your users your app's MCP address. npx buildbase agent install --url <your MCP address> prints the setup for each AI tool, Gemini CLI included, with your address filled in.

  1. In Gemini CLI, your user adds a remote MCP server: gemini mcp add, or settings.json. They use your app's MCP address, https://app.example.com/api/mcp in these examples, or the command or config below.
  2. Sign in: run /mcp auth acme. Gemini CLI finds the sign-in from the server's first answer and opens a browser. It needs a browser on the same machine, so it does not sign in over SSH without port forwarding.
  3. Your hosted sign-in page opens. Your user signs in and approves the consent screen, which names Gemini CLI and what it asks for.
  4. Gemini CLI can now use your tools, as that user and with that user's permissions.
terminalBash
gemini mcp add --transport http acme https://app.example.com/api/mcp
settings.jsonJSON
{
  "mcpServers": {
    "acme": {
      "httpUrl": "https://app.example.com/api/mcp"
    }
  }
}

The plan it needs

Gemini CLI's MCP page does not name a plan. Its documentation says Antigravity CLI replaced it on 18 June 2026 for unpaid and Google One users.

How sign-in looks in Gemini CLI

Gemini CLI finds the sign-in from the server's first answer and opens a browser. It needs a browser on the same machine, so it does not sign in over SSH without port forwarding.

What is different in Gemini CLI

  • Gemini CLI runs on your user's own computer, so it reaches any address that computer can, including a development server while you test.
  • Your server's built-in tools are read-only until you allow changes. includeTools and excludeTools choose the tools (excludeTools wins). trust: true skips every confirmation, so leave it off for a server that can change things.

If it does not work in Gemini CLI

Gemini CLI cannot register or sign in
Turn on Let agents register themselves, with a base client, in the console under Authentication, then Agent readiness. Gemini CLI registers its own client the first time it connects.
Gemini CLI does not connect
/mcp and gemini mcp list show each server's state; --debug opens the debug console. In settings.json, httpUrl is Streamable HTTP and url is the older SSE transport.
Gemini CLI needs to sign in again
Run /mcp auth <name>.
Check your server the way Gemini CLI will
Run npx buildbase agent doctor --url <your MCP address>. It walks the first contact a client makes (the sign-in challenge, the metadata, redirects, a stable tool list) and exits 1 when a check fails.

Frequently Asked Questions

Which Gemini CLI plan do your users need?

Gemini CLI's MCP page does not name a plan. Its documentation says Antigravity CLI replaced it on 18 June 2026 for unpaid and Google One users.

Does my MCP server need to be public for Gemini CLI?

No. Gemini CLI runs on your user's computer, so it reaches what that computer reaches. Your users still need an address they can open, and production should be public HTTPS.

Can Gemini CLI change data in my app?

Only through tools you allow, and only as the user who connected it. The built-in tools are read-only by default, and a tool that needs extra scopes is hidden until the user grants them.

Where next