Ask the Agent about images and screenshots

Ask the Agent about images and screenshots

The Slite Agent can now understand images. Paste a screenshot or upload a picture straight into your prompt, and the Agent will read what's in it alongside your question and your connected sources.
This is useful whenever something is easier to show than to type out: an error on screen, a design, a chart, or a photo of a whiteboard.

How to add an image

    Open the Slite Agent.
    Add your image to the prompt box:
  • Paste it — copy a screenshot or image to your clipboard, click into the prompt box and press Cmd + V (Mac) or Ctrl + V (Windows).
  • Upload it — click the image button in the prompt box toolbar (next to the calendar icon) and choose a picture from your computer.
    Type your question about the image and send it.
The Agent will answer using what it sees in the image, together with the knowledge in your Slite docs and connected data sources.
Tip: Be specific about what you want from the image. "What's causing this error?" or "Does this mockup follow our brand guidelines?" will get a sharper answer than "What's this?".

What you can use it for

Troubleshooting
  • Paste a screenshot of an error message and ask what it means and how to fix it.
  • Share a screenshot of a settings page and ask which option you need to change.
Design and content reviews
  • Upload a mockup and check it against the UX, UI or brand guidelines documented in Slite.
  • Ask the Agent to review the copy in a screenshot against your tone of voice guide.
Understanding visual information
  • Drop in a chart or dashboard screenshot and ask for the key takeaways.
  • Upload a photo of a whiteboard or a diagram and ask for a written summary or next steps.
Saving time on retyping
  • Paste a snippet of a table, a slide or a document instead of retyping it, then ask the Agent to summarise, reformat or compare it with your docs.

Combining images with your knowledge

Because the Agent reads the image and searches your sources at the same time, it can connect the two. For example:
  • "Here's a screenshot of a customer's error — have we documented a fix for this?"
  • "Does this landing page design match our brand guidelines?"
  • "Which of our processes covers the steps shown in this diagram?"
The Agent will cite the docs and sources it used in its answer, so you can check where the information came from.

Related articles

  •  Agent Data Sources⁠ 
  •  Build reusable workflows with Assistants⁠ 
  •  Use the Agent in your browser (Chrome Extension)⁠