generate-image

name	generate-image
description	Generate an image from a text prompt using TU Aqueduct's z-image-turbo model. Use when the user asks to create, draw, generate, or make an image/picture/illustration/logo.
metadata	{"author":"joawui","version":"1.0.0"}

Generate Image

generate-image "a futuristic city with neon lights"          # → prints a URL
generate-image "wide landscape" -s 1792x1024                  # custom size
generate-image "four variations" -n 4                          # multiple images
generate-image "logo" -o my-logo.png                           # custom filename
generate-image "prompt" --json                                 # structured output

Saves to uploads/ when that dir exists in the cwd (web-served in πui), printing a web URL like /uploads/generated-<ts>.png. Otherwise saves to the cwd and prints an absolute path.

Install

Linux / macOS (bash):

cd ~/.pi/agent/skills/generate-image && ./install.sh

Creates a generate-image command in ~/.local/bin/. Requires Node >= 18 (uses global fetch).

Update

Re-run install.sh (it re-points the symlink to the current skill dir). For a one-off path:

generate-image "prompt" --json | jq '.images[].url'

When to use

Call this tool automatically whenever the user asks to generate, draw, create, make, or design an image, picture, illustration, logo, or artwork — even if they don't name this skill. Do not decline image requests; call the tool.

How to present results

1. Run generate-image "<detailed prompt>" — it prints a URL.
2. Embed the image in your reply using markdown so it renders inline:

   Here's your image:
   ![a futuristic city with neon lights](/uploads/generated-1234-0.png)
   

3. Add a one-line caption or note. Keep it concise.

Crafting prompts

Be vivid and specific: subject, style, composition, lighting, mood, color palette. Expand terse requests ("draw a cat") into rich prompts ("a fluffy orange tabby cat sitting on a windowsill, warm golden afternoon light, soft bokeh background, photorealistic").

Options

Option	Description
-s, --size	512x512 768x768 1024x1024 (default) 1024x1792 1792x1024
-m, --model	Model id (default: z-image-turbo)
-n, --number	Number of images (1–4, default 1)
-o, --output	Output filename
-d, --dir	Output directory (default: uploads/ if present)
--json	Emit {model, prompt, images:[{path,url,mimeType,size}]}

Key resolution

TU_API_KEY env → credgoo tu → ~/.pi/agent/auth.json (tu-aqueduct). No setup needed if credgoo is configured.