coloring-page – OpenClaw Skill

---

name: coloring-page description: Turn an uploaded photo into a printable black-and-white coloring page. metadata: clawdbot: config: requiredEnv: - GEMINI_API_KEY

coloring-page

Create a printable black-and-white outline coloring page from a photo.

This skill is designed to be used conversationally:

• You upload an image
• You say: “create a coloring page”
• The assistant runs this skill and sends back the generated PNG

Under the hood, this uses the Nano Banana Pro (Gemini 3 Pro Image) image model.

Requirements

• GEMINI_API_KEY set (recommended in ~/.clawdbot/.env)
• uv available (used by the underlying nano-banana-pro skill)

How the assistant should use this

When a user message includes:

• an attached image (jpg/png/webp)
• and the user asks for a “coloring page”

Run:

• bin/coloring-page --in <path-to-uploaded-image> [--out <output.png>] [--resolution 1K|2K|4K]

Then send the output image back to the user.

CLI

Basic

coloring-page --in photo.jpg

Choose output name

coloring-page --in photo.jpg --out coloring.png

Resolution

coloring-page --in photo.jpg --resolution 2K

Notes

• Input must be a raster image (.jpg, .png, .webp).
• Output is a PNG coloring page on a white background.