Agent Guidebook
Generate Images

Steamship agents are multi-modal: they can send and receive text, images, audio, and video.

This chapter shows you how to generate (or locate) and return images from your agents:

It also shows how to build new image Tools that wrap these tools with your own prompts or logic:

What you can build with this chapter

This chapter will let you build:

  • An agent that can fetch images from the web for you (with Google Image Search)
  • A chatbot that can generate DALL-E or Stable Diffusion images for your friends in a chat room

You can combine these skills with personalities and moods to:

  • Create characters that can show you their imaginary world
  • Create characters whose image generation style is influenced by the mood they're in

You can combine these skills with Steamship's Key-Value storage (opens in a new tab) to:

  • Create characters that remember where the are, and show you photos from that place
  • Create bots that let you and your friends record prompt macros and then use them to generate images