| name | long-description-generation |
| description | How to write a text-equivalent long description (alt text) for a complex image, such as an Azure architecture diagram, decision tree, process flow, or detailed screenshot. USE when generating, reviewing, or improving the alternative text for a complex image so users utilizing assistive-technology can reconstruct it. DO NOT USE for simple decorative images, for icons on their own, or for non-image content. |
| compatibility | Requires the ability to view the image (a multimodal model with the image in context). |
| argument-hint | Attach or point to an existing PNG or SVG image to generate alt text for. |
| disable-model-invocation | false |
| license | MIT |
| user-invocable | true |
| metadata | {"author":"Azure Patterns & Practices"} |
Long description generation
This skill is guidance on how to write a text equivalent for a complex image, usually an Azure architecture diagram. Users of assistive technology, like screen readers, read your long description and build a mental image of the art. You generate a text equivalent to the image.
You act as a web accessibility expert when you apply this skill. The goal is a description precise enough that someone who has never seen the image could reconstruct it fairly well in their mind.
Prerequisite: you must be able to see the image
You need the image in your context to apply this skill. The image is usually one of:
- a cloud architecture diagram of a workload running in Azure
- a screenshot that has important details in it
- a decision tree
- a process flow diagram
If no image is available, stop and ask for one before continuing.
File types
This skill supports raster and vector diagram formats, primarily PNG and SVG. If you're handed a format you can't render or interpret as an image, refuse to operate on it and ask for a different file.
Requirements for the long description
Generate the equivalent alternative text for the image, and adhere to all of the following requirements.
- The text must be no less than 300 characters long.
- The text must be no more than 1000 characters long.
- The text must be in en-US.
- The text must not contain bullet points or ordered lists.
- The text must be one or two paragraphs long.
- Describe the image so that someone reading the text could replicate the image fairly well without having ever seen it.
- Focus on the information the image provides to its consumer.
- Use strategic summarization to group similar concepts, taking cues from any grouping present in the image.
- Be precise and concise.
- Don't describe things that aren't in the image. Describe only the image you were provided.
- Use positional terms and phrases like "above," "below," or "to the left of" to help the reader understand the layout.
- If the image contains arrows or other connectors, discuss those connections and any relationship information they convey.
- Don't describe icons, such as Azure service icons. Address those icons as named components by using the naming in the image. If naming cues aren't available, don't invent a name for the component. Use a generic term for it.
- If there's a legend, describe its contents and use its terms as part of the description.
- If there's a logical flow, such as numbering or a connected chain of arrows, follow that flow when describing the image.
- Assume the reader has additional text that sets the image in a larger context. Don't guess what that larger context is. Focus on describing the image.
