Generate Image with Gemini

Generate images from text prompts using Google's Gemini API.

Arguments

<input_arguments> $ARGUMENTS </input_arguments>

Format: <prompt> <output-path> [options]

Options:

• --aspect - Aspect ratio: 1:1, 16:9, 9:16, 3:2, 2:3, 4:3, 3:4, 4:5, 5:4, 21:9
• --size - Resolution: 1K, 2K, 4K

Examples:

/majestic-creative:generate-image "A cat in space" cat.jpg
/majestic-creative:generate-image "Epic landscape" landscape.jpg --aspect 16:9
/majestic-creative:generate-image "Product mockup" product.jpg --aspect 1:1 --size 2K

Prerequisites

• GEMINI_API_KEY environment variable must be set
• Python 3.10+ with google-genai and Pillow packages

Process

Step 1: Validate Environment

python -c "import os; print('ok' if os.environ.get('GEMINI_API_KEY') else 'missing')"

If missing, inform user:

GEMINI_API_KEY not set. Get one at https://aistudio.google.com/apikey

Step 2: Parse Arguments

Extract from $ARGUMENTS:

• Prompt: Text in quotes or first argument
• Output path: Second argument (ensure .jpg extension)
• Aspect ratio: --aspect value (optional)
• Size: --size value (optional)

Step 3: Generate Image

Use the bundled script (located at skills/gemini-image-coder/scripts/generate_image.py in the plugin):

python generate_image.py "<prompt>" "<output-path>" [--aspect <ratio>] [--size <size>]

Replace placeholders with actual values from the parsed arguments.

Step 4: Verify and Report

Check that the output file was created:

ls -la <output-path>

Report success with file path and size.

Error Handling

Rate Limiting

If you get "API returned no content", wait 2-3 seconds and retry.

Content Policy

If prompt is rejected, suggest modifying the prompt to be more generic.

Multiple Images

When generating multiple images, add 2-second delays between requests:

sleep 2

Output Format

**Generated:** `<output-path>`
**Size:** <file-size>
**Prompt:** "<prompt>"

[Show the image if in a context that supports it]