Use third-party coding agents in Jira automation
This experience is currently in beta.
Automate your development workflow by running third-party coding agents directly in Jira.
Trigger coding agents based on events like new work items or field updates to autonomously handle tasks such as feature flag cleanups and PR creation for bug fixes.
Every action is tracked with robust automation audit logs for full visibility.
Before you begin
AI features must be enabled for your Jira space.
If you haven’t already, you’ll need to set up your preferred coding agent.
To get started with:
GitHub Copilot in Jira, view the app on Marketplace
Cursor in Jira, view the app on Marketplace
Claude in Jira, view the app on Marketplace
Bitbucket Cloud or GitHub Cloud is connected to your Jira site. How to connect a repository
Test your coding agent on a work item before setting up an automation flow. Trigger a coding task by assigning or @mentioning the agent to check it’s authorized. Read more about working with AI agents
To create and manage automations:
At the site-level, you must have Administer Jira permissions.
At the space-level, you must have Administer space permissions in a company-managed space, or Administrator permissions in a team-managed space.
Create an automation rule
This beta release only supports work item and field-based triggers. If you set up a scheduled trigger, we’ll let you know it’s not yet available.
To set up a coding agent in an automation:
If you’re not already there, navigate to Jira using the App switcher ()
Next to your space’s name in the sidebar, select More actions (), then Space settings.
Select Automation.
Or, select Automation from the top right navigation in your space.
Select Create Automation.
In the automation flow canvas, choose either Work item created, or a field-based trigger.
For the next step, add an Action, then search for and select your preferred coding agent.
After you add the coding agent, you can include a custom prompt to describe the task and provide a link to your repository.
Select Save and enable, then add a name for the flow.
Select Turn on flow to activate.
Review and monitor your flow
Once the automation has successfully run, you can review the pull request directly in the repository. Or, open the work item, then from the GitHub Copilot agent or Agents section, select the associated repository link.
View the audit log
Each time a flow is triggered, it saves a log of data that you can audit to see when the flow was triggered, it’s status, and the details of each step the flow attempted to perform. This can be really helpful for debugging and iteration.
Read about how you can monitor automation activity
Was this helpful?