Use casesWorks with

Connect Gemini CLI to your team's memory

Gemini CLI connects to Context Engine with one gemini mcp add command, then /mcp auth and a sign-in in the browser, so it can read and save to the same memory as your team and its other AI tools.

What it does

Gemini CLI starts every session without your team’s context. Connected to an engine, it can search by meaning across the engines the connection reaches, read a memory with its history, and save decisions, notes and tasks that your colleagues and their Claude, ChatGPT or Cursor read too.

Connect it

  1. Sign up at app.contextengine.com and make an engine.
  2. Open AI tools, then Connect an AI tool, and choose Gemini CLI under coding tools. Under Change… pick its engines, each at Can read or Can read and save.
  3. Run the command it shows in your terminal: gemini mcp add --scope user --transport http ce-acme <your connection's address>, with the server named after the connection.
  4. Start gemini in a folder you trust, type /mcp auth ce-acme and say yes when it asks.
  5. Your browser opens. Click Allow.

The screen turns Signed in green when Gemini CLI first reaches the connection, then First memory when it saves something.

No browser where it runs?

Choose Use a key instead under the steps. It makes a personal key named after the app and shows it once, inside a single gemini mcp add command that sends it as an Authorization header to the shared https://app.contextengine.com/mcp address.

Who can see what it saves

Everyone in the engine sees what it saves as “Gemini CLI” followed by whose it is. You can pause or disconnect it from AI tools, and every change records who made it, with which tool, and what it was before.

Questions

Why does Gemini CLI not show the Context Engine server?

Gemini CLI switches MCP servers off in folders you have not trusted. Start gemini in a folder you trust, then run /mcp auth with the server's name.

Why add it with --scope user?

Without it, Gemini CLI adds the server only for the folder you ran the command in. --scope user makes it available wherever you start gemini.

Can I use Gemini CLI on a server with no browser?

Yes. Choose Use a key instead under the steps. You get one gemini mcp add command with the key already in its header.

In the help