Intelligence from Multiple Agents

[Balaji Muniraja]

Hello Community Members,

We are developing agents to streamline Agile Ceremonies, for e.g. Backlog Grooming can be enhanced through Story Readiness Agent. This Agent Calculates a Score called "Readiness_Index" based on the team specific Definition of Ready conditions. We have developed this and working fine for us.

In a similar way, we can develop Agent to see if Retros are done effectively or not. Like I can give instructions to find, Scrum Master leveraged Burn Downs, Discussed Defects, Spill Overs, etc., and give a Retro_Effectiveness_Score.

Wanted to know if ROVO can co-relate the outcomes of Backlog Grooming to Demo to Retro using a common chat or do we need to build an agent for this?

If we need to build an agent for this, how to go about it ? Please someone can share the steps for the same.

Thanks and Regards

Balaji.M

[Dr Valeri Colon _Connect Centric_]

@Balaji Muniraja Rovo can’t “merge intelligence” across multiple agents automatically today. Each agent run is scoped to its own instructions, knowledge, and the current chat context, so one agent won’t reliably consume another agent’s scores unless you explicitly persist them somewhere.

Best approach: build one orchestration agent (using Scenarios) that covers Grooming → Demo → Retro, and have each scenario write the computed scores back to Jira/Confluence (e.g., custom fields like Readiness_Index, Retro_Effectiveness_Score, or a Confluence summary page). Then the next scenario can read those stored values and correlate them. Without storing outputs in Jira/Confluence, long chats will drop earlier details and correlation will be inconsistent.

P.S. Everytime I say "Grooming" now I think of @Alex Ortiz 

[Alex Ortiz]

@Dr Valeri Colon _Connect Centric_ grooming user stories is a thing!

[Dr Valeri Colon _Connect Centric_]

But you brought new meaning to it @Alex Ortiz 

[Matteo Vecchiato]

Hi  @Balaji Muniraja ,

As far as I know, Rovo agents can collaborate directly between them. What I can suggest is to automatically save the scores from the two agents to work items fields and then generate a third Rovo agents that read thos information.

Hope it helps