What it does
Writes one careful prompt from your request (subject, style, composition, palette, what to avoid) and makes the picture, which lands in your Files. It keeps your house style in mind. It cannot see what it made, so it tells you what it asked for, and it never copies a brand's assets or draws a real person.
Ask it things like
- “A watercolour of our street in autumn.”
- “An icon for my podcast.”
- “Same picture, wider.”
What it can use
6 tools it uses without asking. You see the whole list, tool by tool, before you add it.
- Image1 without asking
Makes images with the Image plugin.
The tools
image.generate - Memory5 without asking
Remembers what you tell it, and forgets it when you ask.
The tools
memory.forgetmemory.get_preferencesmemory.notememory.recallmemory.remember_preference
What it knows how to do
- Writing an image promptHow one request becomes one image prompt: subject, style, composition, palette, what to avoid; references; no text unless asked; no brands or real people. Read it before making an image.
What it needs
- Image plugin, 0.1.3 or newer. Installed on the way when you add the agent.
- An image account, which buddi asks you to pick.
How to add it
Agents → Catalogue → Illustrator. Its page shows the tools, the picks and the missions, and adds it on your yes.
“Add the Illustrator from the catalogue.” He brings up the same page for your yes.
buddi agents add illustrator
Version 1.0.1
Its image-prompt skill reads in plain words.