What is this project

This is a project to take business requirements from Confluence and use GPT-3 to convert the requirements to JIRA tickets for developers to work on.

How is this done?

This is done by the following flow:

• Pulls the contents of a page from Confluence
• Using pre-defined prompts, we ask GPT-3 to generate the following information for us:
  • Problem Statement
  • User Story (as a XXX)
  • Frontend tasks
  • Backend tasks
  • QA testing tasks
• We place the generated information into a JIRA template generated using the Atlassian ADF Builder.
• We create a JIRA ticket using the JIRA API.
• We link the Confluence page to the JIRA ticket so developers can go back to requirements and vice versa.

The purpose of this project

This project was created as part of a hackathon. This proof of concept was completed in ~4 hours and work hasn't been done on it since. This code can still be used if anyone wants to expand on this idea.

Architecture

Limitations

There are several notable limitations considering the project was developed in a few hours:

• The flow is limited to 1 JIRA ticket per Confluence page.
• There is currently a bug where the Confluence page is added via the Remote Link API but the link doesn't display in JIRA.
• The Confluence page ID is hardcoded, so code will need to be added to make it variable.

How to run

To run, just clone the project then run python main.py.

You must have the following environment variables before running:

• OPEN_API_KEY: OpenAI API key that you can retrieve in the API Keys page.
• ATLASSIAN_API_KEY: Atlassian API key to make Confluence and JIRA API calls.
• MAIN_DOMAIN: The main domain name of your Atlassian pages. (https://.atlassian.net/).
• ATLASSIAN_EMAIL: Your Atlassian e-mail.