Kolbo.AIKolbo.AI Docs

Prompt Copilot

An in-app assistant that takes instant actions inside your tools, keeps your selected model, understands @Visual-DNA and

The Kolbo assistant no longer just answers questions. The Prompt Copilot takes action inside your tools, remembers your choices, and speaks the language of your creative work.

Overview

Ask the Copilot to make something and it does more than describe how. It generates the result and attaches it straight to the right tool, ready for you to keep working. It respects the model you picked, understands your saved characters and styles by name, shapes prompts from plain-language style requests, and points you to the exact lesson when you could learn a feature faster by watching it.

How It Works

Open the chat and ask the Copilot to make something. It interprets the request, takes the action inside the matching tool, and hands you the output where you need it.

Instant Image Actions

Ask for an image and the Copilot generates the result and attaches it to the right tool. This means:

  • No copy and paste between chat and your canvas
  • The output lands exactly where you need it
  • One request, ready to refine immediately

It Keeps Your Selected Model

When you have already picked a model, the Copilot uses it. Ask for an image or an edit and it respects your choice instead of falling back to a default:

  • The model you selected is the model it uses
  • No surprise switches mid-task
  • Consistent results across a session

It Understands Your Creative References

Kolbo lets you save your characters and looks as reusable references, and the Copilot reads them too:

  • @Visual-DNA mentions for your saved characters and identities
  • #moodboard mentions for your saved styles and aesthetics

Bring a character or a mood into the conversation by name and the Copilot knows exactly what you mean, then applies it to what you are making.

Style Requests in Plain Language

Want something to feel a certain way? Just say it. Ask for anime, 2D, watercolor, or any other look and the Copilot shapes the prompt to match:

  • "Make it anime"
  • "Give it a 2D illustrated feel"
  • "Turn this into a watercolor"

No prompt engineering required. Describe the vibe and the Copilot handles the rest.

Learning Hub Recommendations

When the Copilot senses you could learn a feature faster by watching it, it points you to the right lesson in the Learning Hub. You get unstuck without searching, learn the tool you are already using, and move from question to mastery in one step.

Best For

  • New users: get instant results and the right lesson at the right moment
  • Power creators: reference your saved characters and moodboards by name and let the Copilot act inside your tools without breaking your flow
  • Visual storytellers: describe the style you want in plain words and get prompts shaped to match

Tips

  1. Pick your model first - the Copilot will keep using it
  2. Reference saved characters with @Visual-DNA and styles with #moodboard
  3. Describe the look you want in plain language instead of engineering a prompt
  4. Follow the lesson suggestions to learn a tool while you use it
  5. Ask it to generate directly so the result lands in the right tool