Issue Summary

Currently, Rovo agents are limited to interacting with users and cannot communicate directly with each other. This restriction makes it difficult to build advanced, multi-agent workflows where the output of one agent seamlessly becomes the input for another. While automations can connect agents to some extent, they are cumbersome and do not allow for true orchestration—especially since agent actions require user confirmation and cannot be triggered autonomously.

Suggestion:

I propose enhancing Rovo to support native agent-to-agent communication, allowing agents to delegate tasks, share data, and trigger each other’s actions within a single workflow, without relying on external automations or manual user intervention. This would enable scenarios such as a main agent orchestrating specialized agents (e.g., for story progress, blockers, dependencies, and reporting) to deliver comprehensive, integrated responses to user queries like “complete project status report.” Such a feature would greatly improve workflow automation, reduce manual steps, and unlock the full potential of multi-agent collaboration in the Atlassian ecosystem.