Tabnine AI agents generate, validate code for Jira issues

AI coding assistant Tabnine has added two AI agents that integrate with the Atlassian Jira project management platform, one for generating code from requirements outlined in Jira issues and one for validating code for those requirements.

Unveiled September 24, the two agents include the Jira Implementation Agent and the Jira Validation Agent. With one click, developers can implement a Jira issue, whether a story, bug, task, or subtask, and the implementation will generate code from requirements in that issue, Tabnine said. Afterward, developers can use the validation agent to ensure that selected code, whether human- or AI-generated, meets the specifications of an issue in Jira. Tabnine offers instant feedback and code suggestions if adjustments are necessary.

The two agents support what Tabnine described as its AI programming assistant’s ability to automate the creation of as much as 50% of code and artifacts for developers. Privacy and protection from legal risk are offered by the company, with Tabnine AI agents respecting the company’s zero data retention policy for information exposed through Jira. Instructions on getting started with Tabnine can be found at tabnine.com. A demo of the two agents can be found on YouTube.

According to Tabnine, the capabilities of the new Jira agents include:

• One-click code generation, generating code for Jira issue requirements using the Jira Implementation Agent.
• AI-driven code validation using the Jira Validation Agent.
• Implement parent issues like entire Jira stories, bugs, and tasks directly through both agents.
• Contextual AI code suggestions also are offered through both agents.
• Enterprise-ready configuration, with admin-controlled deployment, also leverages both agents.