Nano Banana Pro — Gemini 3 Pro Image Generation

Script: ~/clawd/skills/nano-banana-pro/scripts/generate_image.py Runner: uv (installed at ~/.local/bin/uv) Requires: GEMINI_API_KEY env var from https://ai.google.dev/

When to Use

• User wants to create/generate images from a text description
• User wants to edit/modify an existing image
• Multi-image composition (up to 14 inputs)
• Cost-sensitive image generation (cheaper than gpt-image)

When NOT to Use

• Screenshots/screen capture → use browser tools
• Photo analysis without generation → just describe the image
• Video generation → use video-generation skill

Generate an Image

uv run ~/clawd/skills/nano-banana-pro/scripts/generate_image.py \
  --prompt "your image description" \
  --filename "output.png" \
  --resolution 1K

Edit an Existing Image

uv run ~/clawd/skills/nano-banana-pro/scripts/generate_image.py \
  --prompt "edit instructions" \
  --filename "output.png" \
  -i "/path/to/input.png" \
  --resolution 2K

Multi-Image Composition (up to 14 images)

uv run ~/clawd/skills/nano-banana-pro/scripts/generate_image.py \
  --prompt "combine these into one scene" \
  --filename "output.png" \
  -i img1.png -i img2.png -i img3.png

Control Aspect Ratio

uv run ~/clawd/skills/nano-banana-pro/scripts/generate_image.py \
  --prompt "portrait photo" \
  --filename "output.png" \
  --aspect-ratio 9:16

Options

• --resolution: 1K (default), 2K, 4K
• --aspect-ratio / -a: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9
  • Without this flag, the model picks freely
  • Use for avatars, profile pics, or consistent batch generation

Output

• The script saves the image and prints the path
• Use MEDIA:/path/to/output.png in response to send as attachment
• Use timestamps in filenames: yyyy-mm-dd-hh-mm-ss-name.png

Pitfalls

• ⚠️ Requires GEMINI_API_KEY — check if set before running
• Do NOT read the image back after generating — just report the saved path
• For Hermes: include MEDIA:/absolute/path in response to auto-attach