TechPrufer
Get started
Atlassian
Atlassian logoAtlassian · Feature trail

Agentic Pipelines for Maintenance Tasks

2 posts · Mar 2026 – Sep 2026

What this trail covers

Atlassian is developing Agentic Pipelines to automate recurring maintenance tasks, starting with feature flag cleanup. This capability aims to reduce engineering overhead by using AI agents to prepare code changes, such as removing stale feature flags, and opening pull requests for human review. The system emphasizes a 'thin prompt, rich skill' architecture, robust validation, and scoped permissions to ensure safety and efficiency. This pattern is being explored for other maintenance tasks like turning Loom recordings into dev-ready work items in Jira, capturing technical context for bug reports, and suggesting Jira updates from meeting discussions.

All Atlassian trails
  1. How we automated feature-flag cleanup with Agentic PipelinesThis post introduces Agentic Pipelines as a new capability for automating feature flag cleanup. It details a two-part workflow: a dispatcher (Rovo Studio Automation rule) that finds stale flag-cleanup tickets in Jira and triggers a coding agent pipeline, and the coding agent itself which searches the codebase, inlines final behavior, removes dead code, runs checks, and opens a pull request. The post highlights the importance of a 'skill' file for domain logic, verification before code changes, and gating PRs with checks. It also discusses lessons learned and potential applications for other maintenance tasks.
  2. Turn Loom meetings and bug reports into Jira work with AIThis post introduces new AI workflows for Loom that integrate with Jira and Rovo. The enhanced bug reporting mode automatically captures technical context (browser, device, network, console logs) and turns it into a dev-ready Jira work item. AI-suggested Jira work item updates based on Loom meeting recordings and recaps are also introduced to capture decisions and action items. The system leverages Loom's technical and conversational context to drive end-to-end bug fixes, incident response, and task management in Jira, Confluence, and Rovo.