GitLab has introduced an event-driven Work item created trigger that fires the instant a new work item appears in a project, turning routine issue triage and assignment from a manual daily chore into automation that completes in seconds.
The Cost of Manual Triage at Scale
Assigning incoming work manually involves dozens of small decisions every day: for each new issue, someone has to pause, check who has capacity, weigh current workloads against what's coming in, and decide where the item should go, all while accounting for meetings, breaks, and time off. That process works fine at small scale but breaks down as volume grows, leading to delayed triage, uneven workload distribution, and team leads spending time on routing instead of higher-value work.
Why This Wasn't Possible Before
Previously, every GitLab flow required a manual action, a mention, an assignment, or a similar UI event, to start running, meaning even a well-designed assignment flow still waited on a person to trigger it. The new trigger closes that gap: it fires automatically the moment a work item is created, with no manual handoff required, letting a flow spring into action on its own to handle the routing.
How the Demo Flow Actually Works
GitLab built a two-agent flow called Work item assigner to demonstrate the trigger. The first agent uses GitLab Orbit, a lifecycle context graph for software engineering, to pull each team member's current count of open work items across the top-level group. The second agent then identifies whoever carries the lightest current load and assigns the new item to them directly. Because the assignment logic lives entirely inside the flow itself, the trigger alone is enough to set the whole process in motion; in GitLab's demonstration, the flow correctly identified and assigned a new issue to the least-loaded teammate within seconds of its creation.
Where the Approach Could Go Next
GitLab suggests further extending the flow with Model Context Protocol connections to HR or PTO systems and to team calendars, so the assigning agent could factor in availability and time off alongside current workload, moving the automation closer to how a thoughtful human team lead would actually make the call.
The Bigger Shift
The change reflects a broader move in GitLab's Duo Agent Platform toward event-driven automation rather than automation that still waits for a person to kick it off, freeing team leads from a routine daily decision that an agent can make with the same information a human would use, so people can focus attention on work that genuinely requires judgment.