What is the challenge? |
When defining AI Instructions, the current feature does not allow the instruction to reference images or other Aha records |
|---|---|
What is the impact? |
It is difficult to provide the Aha Agent instructions to follow in the context of prototypes as I can not reference design language in the form of images |
Describe your idea |
Expand the instruction feature to a. Allow the insertion of images b. reference other Aha records types/documents |
AI skills are here
Create reusable instructions to inform Elle how you like to complete specific work. Define a skill once and Elle applies it automatically — no need to add it to the chat yourself.
Adding a new use case to this Idea. We are working on a agent that drafts user stories (Requirements). The feature has reference images, screenshots, etc within it and we have asked the agent to inspect the images within the feature and use them within the requirement if appropriate. This gap in feature does not allow us to full implement this feature within the agent
This is the primary reason I am unable to utilize this feature. As with other AI solutions, training a gpt/project/skill for reuse, often you are uploading documents -- having to figure out how to translate, if even possible, to a text box makes it a nightmare, especially with maintenance. Therefore, the value out of Aha! AI is very limited and we pay for another solution, and then connected Aha! to it. We would have preferred to remain within Aha!, but not an option.