All products

Write exhibition text with AI

Turn completed research into visitor copy with a clear opening, concise labels and a source attached to each interpretation.

To write exhibition text with AI, start with a question a visitor can answer by looking. A station clock has two sets of screw holes. An inventory records that the face was replaced, but says nothing about who fitted it. Give research-to-curatorial-copy the image, the inventory entry and your notes. It can introduce the visible detail, explain the documented change and leave the fitter’s identity open.

The output gives each part of the text a job. A short opening directs the visitor’s attention. A longer passage connects the observation to the exhibition’s subject. A source entry lets an editor trace that connection before putting the words on a wall or a website.

What it does

  • Find two to four themes that recur across the material you supply.
  • Draft an object label around a detail visible in the photograph.
  • Give a group of exhibits a heading and a short introduction.
  • Pair a date or interpretation with the archive entry that supports it.
  • Flag an unresolved attribution for editorial review.

Install and use

  1. Install the skill in your agent and supply research notes, object descriptions and their existing sources.
  2. Choose a visitor entry point and ask for a heading, summary, main text and evidence for each unit.
  3. Read the proposed copy beside the material and check which sentences still need a qualification.

FAQ

Source

Made by Kychrona. Original skill: research-to-curatorial-copy, published on GitHub under the MIT licence.

View on GitHub