gemini

---
name: gemini
description: Generate an image using Google Gemini via Chrome browser automation. Use when the user says "generate an image with Gemini", "use Gemini to make", "Gemini image", "make an image in Gemini", "create with Gemini", or wants to use Google's Gemini web UI to generate images. Takes an image prompt as the argument.
user-invocable: true
allowed-tools:
  - mcp__claude-in-chrome__tabs_context_mcp
  - mcp__claude-in-chrome__tabs_create_mcp
  - mcp__claude-in-chrome__navigate
  - mcp__claude-in-chrome__read_page
  - mcp__claude-in-chrome__find
  - mcp__claude-in-chrome__form_input
  - mcp__claude-in-chrome__computer
  - mcp__claude-in-chrome__get_page_text
  - mcp__claude-in-chrome__javascript_tool
  - Bash
  - Read
---

# Gemini — Generate Images via Chrome

Arguments passed: `$ARGUMENTS`

## Argument Handling

- `$ARGUMENTS` is the image prompt (e.g., "a cat riding a skateboard in a cyberpunk city")
- If no arguments provided, ASK the user what image they want to generate

## Workflow

### Step 1: Get Browser Context

Call `mcp__claude-in-chrome__tabs_context_mcp` to see current tabs.

- If Gemini is already open in a tab, reuse it
- Otherwise, create a new tab

### Step 2: Navigate to Gemini

Navigate to `https://gemini.google.com/app`

- Wait for the page to load
- If the user is not signed in, tell them: "You need to be signed into your Google account in Chrome. Please sign in and try again."

### Step 3: Enter the Prompt

1. Read the page to find the input/textarea for the prompt
2. Type the image generation prompt. Prefix with "Generate an image of: " if the user's prompt doesn't already indicate image generation
3. Submit the prompt (press Enter or click the send button)

### Step 4: Wait for the Image

1. Wait for Gemini to process — this can take 10-30 seconds
2. Read the page periodically to check if the image has been generated
3. Once the image appears, confirm to the user:
   ```
   Image generated in Gemini. You can view it in your browser tab.
   ```

### Step 5: Download the Image

1. Look for a download button on the generated image and click it
2. If no download button is visible, try right-clicking the image and saving it
3. Check the default downloads directory (`~/Downloads/`) for the newly downloaded file
4. Report the downloaded file path to the user:
   ```
   Image downloaded: ~/Downloads/<filename>
   ```

## Rules

- **Don't trigger alerts or dialogs** in the browser
- **If the page structure is unexpected**, describe what you see and ask the user for guidance rather than clicking blindly
- **If Gemini shows an error** (content policy, rate limit, etc.), report it clearly to the user
- **Keep it simple** — navigate, type, submit, confirm. Don't over-automate.