Tool: generate_image
The generate_image tool creates an image from a text description (prompt) and saves it to media storage. The result can be used as an activity cover.
What it does#
- Accepts a text description of the desired picture
- Deducts 20 ⚡️ energy from the balance
- Sends a request to the image generation service
- Saves the result to media storage (category
support-generated) - Sends the finished image to the chat
How to call it#
Describe the image you need:
- "Generate a picture for a giveaway"
- "Draw an image with confetti and prizes"
- "Create a cover for a movie-themed contest"
- "Generate an image: tropical beach with contest banner"
It's better to phrase the prompt in English - this improves the quality of the result.
Requirements#
- Authorization: required.
- Energy balance: at least 20 ⚡️.
- The generation service must be available - if not, the agent will let you know.
Cost#
Each generation costs 20 ⚡️ energy. Energy is restored daily at 00:00 UTC according to your plan.
If there isn't enough energy, the AI automatically switches to sparks (✨). If there's neither enough energy nor enough sparks, the agent will report an error.
Usage example#
Request: "Draw a cover for a contest with green test tubes and a science lab"
The agent:
- Translates the description into an English prompt
- Deducts 20 ⚡️
- Generates the image
- Sends it to the chat with an rms://... link in media storage
The resulting image can be set as the cover when creating an activity.