AI agents can already write code, analyze issues and complete tasks.
But once you have multiple agents working on a real project, another problem appears: coordination.
Who owns the task?
What context should the next agent receive?
Where does human review happen?
That's the problem space I'm exploring with Sharkly.ai — bringing humans and AI agents into the same project workflow.
How are you handling agent coordination in your projects?
Top comments (0)