# gflow-cli MCP server

> Drive Google Flow's Veo video and Imagen image generation models through an AI client or terminal.

- Repository: https://github.com/ffroliva/gflow-cli
- Page: https://hostedmcp.org/servers/ffroliva/gflow-cli
- Transport: stdio
- Stars: 393
- License: MIT
- Latest version: 0.99.0

## Install

### Claude Code

```sh
claude mcp add gflow-cli -- uvx gflow-cli
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "gflow-cli": {
      "command": "uvx",
      "args": [
        "gflow-cli"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "gflow-cli": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "gflow-cli"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"gflow-cli","type":"stdio","command":"uvx","args":["gflow-cli"]}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "gflow-cli": {
      "command": "uvx",
      "args": [
        "gflow-cli"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add gflow-cli -- uvx gflow-cli
```

### Codex config.toml (`~/.codex/config.toml`)

```toml
[mcp_servers.gflow-cli]
command = "uvx"
args = ["gflow-cli"]
```

### Gemini CLI

```sh
gemini mcp add -s user gflow-cli uvx -- gflow-cli
```

### Gemini settings.json (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "gflow-cli": {
      "command": "uvx",
      "args": [
        "gflow-cli"
      ]
    }
  }
}
```

## Environment variables

| Name | Required | Secret | Description |
|---|---|---|---|
| `GFLOW_CLI_HOME` | no | no | Where gflow keeps its Chrome profile, config and database. Defaults to a per-user directory. |

## FAQ

### How does the tool handle authentication?

It uses a persistent Playwright Chromium profile to handle Google authentication and reCAPTCHA gates.