> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cognee.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Gemini CLI with Cognee Cloud

> Connect Gemini CLI to your Cognee Cloud workspace over MCP so it can remember and recall from your memory.

[Gemini CLI](https://github.com/google-gemini/gemini-cli) supports the [Model Context Protocol](https://modelcontextprotocol.io). Register Cognee as an MCP server and Gemini can `remember`, `recall` and `forget` against your Cognee Cloud workspace. The **Gemini CLI** card on the [Integrations](https://docs.cognee.ai/cognee-cloud/ui/integrations.md#mcp-clients) page shows the same steps with your workspace URL and API key filled in.

## 1. Install uv

Cognee runs through `uvx`, which comes with [uv](https://docs.astral.sh/uv/). There is nothing else to install.

* **macOS:** `brew install uv`
* **macOS/Linux:** `curl -LsSf https://astral.sh/uv/install.sh | sh`
* **Windows (PowerShell):** `powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"`

## 2. Add Cognee to your Gemini settings

Gemini CLI keeps its settings in `~/.gemini/settings.json`. If the file or the `~/.gemini` folder does not exist yet, run `gemini` once to create them, or create the file yourself.

Add the `cognee` entry under `mcpServers` and save. Use your workspace URL and a key from the [API keys](https://docs.cognee.ai/cognee-cloud/ui/api-keys.md) page:

```json theme={null}
{
  "mcpServers": {
    "cognee": {
      "command": "uvx",
      "args": [
        "cognee-mcp@0.5.6",
        "--api-url", "https://your-tenant.aws.cognee.ai",
        "--api-token", "your-api-key"
      ]
    }
  }
}
```

> [!NOTE]
>
> If you already have an `mcpServers` block, add only the `cognee` entry inside it instead of replacing the file.

## 3. Verify

1. Start Gemini in your terminal with `gemini`.
2. Type `/mcp`. You should see **cognee** listed with its tools.
3. Ask *What do you know from cognee?* Gemini should call a **cognee** tool and answer from your memory.

If `/mcp` does not list cognee, Gemini usually cannot find `uvx` on its `PATH`. Run `which uvx` (`Get-Command uvx` on Windows) and use that absolute path as the `command` value.

***

> [card] [Cloud MCP](https://docs.cognee.ai/cognee-cloud/connections/cloud-mcp.md)
>
> How MCP clients connect to Cognee Cloud

> [card] [All agent integrations](https://docs.cognee.ai/cognee-cloud/agent-integrations.md)
>
> Every agent and tool that can use Cognee as memory


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.