GPT for Jira - AI Generated Acceptance Criteria & Test Cases

GPT for Jira is a tool for Agile teams that generates acceptance criteria and test cases based on the user story description.

Our scrum teams were struggling with defining clear and testable user story requirements within the limited time blocks we had with product owners and stakeholders. Often, we’d spend 15 minutes discussing requirements and end up with just a couple of vague sentences for a single story.

So I had a crack at it with GPT-4. It quickly processes the user story description and generates acceptance criteria in about 15 seconds. The coverage is impressive, as it thinks of aspects we might not consider until later in the conversation (if at all). It sometimes suggests criteria we don’t need but being able to quickly remove a couple of points and move on to the next story is so much more efficient.

Integrating this tool into Jira enables you to efficiently review user stories, allowing for the completion of more stories with consistent and clear output in the same timeframe.

Some examples:




Congrats on your launch. Best of luck!

2 Likes

What are the limitations? Seems like it can only work for small amount of text, not tables, or able to skip through links

1 Like

Hi Sharma!

So the app is limited to text input right now. When I first implemented it, there were some issues with the API latency, so the response was pushing Jira’s timeout limit. If I go beyond text, I have to take into consideration the time it takes to parse images/links and then get the GPT-4 output from those. If a user includes many links, it will take a while to parse through them and send the response in time.

However, I’ve got some QoL improvements planned to get around the timeout issue, and add more flexible functionality (including image and link processing). I’m the only engineer on this project so it will take some time to test but I’m going to incrementally roll out some new features over the coming weeks/months.